Characterization of the Microbiome of Freshwater Sponges Undergoing Asexual Reproduction

CHARACTERIZATION OF THE MICROBIOME OF FRESHWATER SPONGES UNDERGOING ASEXUAL REPRODUCTION

A Thesis by TAYLOR A. STROPE

Submitted to the Graduate School Appalachian State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE

August 2020 Department of Biology

CHARACTERIZATION OF THE MICROBIOME OF FRESHWATER SPONGES UNDERGOING ASEXUAL REPRODUCTION

A Thesis by TAYLOR A. STROPE August 2020

APPROVED BY:

Cara Fiore, Ph.D. Chairperson, Thesis Committee

Zack Murrell, Ph.D. Member, Thesis Committee

Cortney Bouldin, Ph.D. Member, Thesis Committee

Cole Easson, Ph.D. Member, Thesis Committee

Mike McKenzie, Ph.D. Dean, Cratis D. Williams School of Graduate Studies

CHARACTERIZATION OF THE MICROBIOME OF FRESHWATER SPONGES UNDERGOING ASEXUAL REPRODUCTION

Taylor Ann Strope B.S., Texas A&M University at Galveston M.S., Appalachian State University Thesis Committee Chairperson: Dr. Cara Fiore

Sponges are sessile filter-feeding organisms that are found in both marine and freshwater habitats and are known to harbor symbiotic microbial communities. There is strong interest in understanding the taxonomic and functional composition of these microbial communities due to their integral roles in sponge ecology, nutrient cycling, and production of natural products. Particularly with freshwater sponges, there is limited work on the functional roles of the sponge microbiome. The goals of the current work were to first understand how the composition of these microbial communities change prior to, and post production of asexual reproductive bodies known as gemmules in freshwater sponges. I also aimed to determine what role, if any, the microbiome may have in the facilitating developmental change of the sponge host.

Preliminary taxonomic profiling by 16S rRNA gene sequencing of sponges and river water indicated the presence of putatively sponge-enriched bacterial taxa with approximately 60% of taxa unique to the sponges. Metatranscriptome analysis of five sponge microbiomes collected prior to gemmule formation (“pre” samples) and three separate sponge microbiomes post gemmule formation (“post” samples) highlighted four key results: 1) taxa within the phyla Bacteroidetes and Proteobacteria (class

iv Alphaproteobacteria) were significantly different between pre and post samples in the gene expression analysis and these taxa were identified as enriched in the sponges based on 16S taxonomic profiling, 2) the presence of eukaryotic-like domains and a type IX secretion system that may be important in symbiosis, 3) multiple functional transcripts were enriched in the post samples (e.g., porin protein, pilus assembly protein) potentially indicative of long-term symbionts that may be present in all life stages of the sponge, 4) multiple taxa within the class Flavobacteriales that decreased in the post samples, potentially indicative of facultative symbionts. This study has implications for better understanding the functional roles of the microbiome in freshwater sponges and draws implications that may be relevant to marine species and other microbial symbiont systems.

v Acknowledgments

I want to acknowledge those who helped fund this research: Office of Student

Research at Appalachian State University, the Appalachian State University Biology

Department, GRAM Award, and an NSF award 1756171 in Biological Oceanography and

Integrated and Organismal Systems. As well as my committee: Dr. Cara Fiore, Dr. Zack

Murrell, Dr. Cort Bouldin, and Dr. Cole Easson. Of course, my labmates in Dr. Cara

Fiore’s lab. In addition to several personal in the Appalachian State University Biology and Chemistry Department including, but not limited to: Dr. Rose, Dr. Venable, Dr.

Creed, Dr. Valley, Dr. Babyak. In addition to in house help, I had assistance from several outside sources such as Dr. April Hill from the University of Richmond, Smithsonian

Tropical Research Institute-specially Dr. Thacker and Dr. Collin, the Marion Freshwater

Mussel Hatchery in Marion, North Carolina, and GENEWIZ Laboratory. Lastly, my father and the rest of my family and friends, I owe them a great deal of appreciation.

vi Table of Contents Abstract……………………………………………………………………………………………………………………iv

Acknowledgments ...... vi

Introduction...... 1

Introduction to Freshwater Sponges ...... 1

Identification of Freshwater Sponges………………………………………………………………………2

Reproduction of Freshwater Sponges with a Focus on Asexual Reproduction…………...4

Sponge Cell Types, Structures, and Development from a Gemmule…………………………...5

Microbiome of Phylum Porifera………………………………………………………………………………9

Methods ...... 15

Field Methods ...... 15

Laboratory Methods……………………………………………………………………………………………..16

Preliminary Microbiome Profiling of 16S rRNA Genes ...... 16

RNA and DNA Extractions and Sequencing for Metatranscriptome……………………...17

Sequence Assembly of Symbiont Metatranscriptome………………………………………….18

Environmental Data………………………………………………………………………………………….24

Results……………………………………………………………………………………………………………………25

Sponge Identification and Preliminary Microbiome Profiling of 16S rRNA Genes ...... 25

Metatranscriptome Results…………………………………………………………………………………..26

Environmental Results…………………………………………………………………………………………30

Discussion……………………………………………………………………………………………………………….31

16S rRNA Microbiome Profiling ...... 31

Metatranscriptome Analysis of Sponge Symbiont Communities……………………………..32

A Potential Framework for Symbiont-Host Interaction in Freshwater Sponges…...…..34

vii Bibliography ...... 41

Tables and Figures…………………………………………………………………………………………………..50

Appendix 1...... 74

Vita…………………………………………………………………………………………………………………………77

viii 1. Introduction

1.1 Introduction to Freshwater Sponges

Sponges are in Phylum Porifera, the most basal clade of metazoans and appear in the fossil record approximately 600 million years ago (Love et al., 2009). The Porifera have proliferated to many diverse forms and lineages in the ocean, while only one major lineage is present in freshwater worldwide: Order Spongillida in Class Demospongiae (Manconi &

Pronzato, 2002). The suborder Spongillida, contains 6 families, 45 genera, and 219 species currently described (Manconi & Pronzato, 2008). In contrast, marine sponges have over 10,000 species belonging to 107 genera within 21 orders (Zhang et al., 2003). Most marine species are within the Class Demospongiae, although sponges in other classes are also represented (i.e., glass sponges, calcareous sponges, and encrusting homoscleromorph sponges).

Freshwater sponges are sessile, filter feeding animals that inhabit a variety of freshwater systems including streams, lakes, rivers and ponds and can be found in both lentic

(still water) and lotic (flowing water) habitats (Manconi & Pronzato, 2008). Sponges attach themselves to various substrata submerged in the water, including rocks, bivalve shells, and man-made debris (Pronzato & Manconi, 2002; Ricciardi & Resiwig, 1992). These organisms also have the ability to tolerate extreme chemico-physical conditions (e.g., desiccation, cold temperatures) (Harrison, 1974; Pronzato & Manconi, 2002). The ability to survive such conditions is in part, due to an asexual life cycle that includes a resting stage called gemmules, which can survive harsh environmental conditions to produce a new sponge when the environment is favorable again. The production of gemmules is characteristic of freshwater sponges, with very few marine sponges carrying this trait. 2

Although freshwater sponges are not as abundant or speciose as marine sponges, they still play an important role in the ecology of freshwater environments. They are prolific filter feeders and play an active role in recycling suspended organic materials that reside in the water column (Manconi & Pronzato, 2008). Many freshwater sponges are also important in primary production and nitrogen cycling with some sponges hosting high numbers of unicellular algal symbionts (Frost, 1978; Frost & Williamson, 1980). Freshwater sponges also host other microbial symbionts that may have important roles in nutrient dynamics and in host function

(e.g., Gaikwad et al., 2016). However, the characterization of the microbiome of freshwater sponges lags behind that of marine sponges (Thomas et al., 2016). Sponges are also an important food source for other animals that inhabit freshwater systems, including insects, waterfowl, fishes, and turtles (Manconi & Pronzato, 2007; Ricciardi & Resiwig, 1992). Thus, there is interest in better understanding their ecological roles and the factors that influence their development and survival. Such factors include environmental parameters as well as associated microbes within the sponge (Gaikwad et al., 2016; Pronzto & Manconi, 2002). While the exact nature of the relationship of the associated microbes within freshwater sponges is not known, most are considered as commensals (similar to marine sponges) (Keller-Costa et al.,

2014; Schmitt et al., 2007; Taylor et al., 2007), where the host may not benefit from harboring the microbes, but it is also not harmed. I use the term microbiome to include the community of associated microbes within the sponge and the term symbiont to refer to sponge-associated microbes regardless of their beneficial, negative, or neutral interaction with the host (de Bary,

1879).

1.2 Identification of Freshwater Sponges

Freshwater sponges are among the least studied faunal group for several reasons

(Ricciardi & Resiwig, 1992); the primary challenge is their highly variable morphologies (e.g.,

3 body shape, dimension, texture, color) between species and even among individuals (Manconi &

Pronzato, 2008). The lack of distinguishable morphological features also makes differentiating between species in the field difficult (Ricciardi & Resiwig, 1992). Furthermore, there are few informative taxonomic keys for freshwater sponges as the taxonomy is not well resolved. Most of the current taxonomic keys lack basic information on the different species, such as ecological variation, environmental factors that can be correlated with individual species, and a range of morphological variation among the different species (Ricciardi & Resiwig, 1992).

The skeletal make up of freshwater sponges, which is used in species identification, is comprised of collagen and spongin fibers embedded with spicules to provide the animal with strength and flexibility (Aguliar-Camacho et al., 2019). Spicules are siliceous, needle-like structures that comprise the mineral skeleton of the sponges and form the structural support for the sponge’s filtration system (Bart et al., 2019). Taxonomy for sponges is based on the morphology (e.g., shape and size) and arrangement of these spicules (Hooper & Van Soest, 2002;

Manconi & Pronzato, 2008; Morrow & Cárdenas, 2015). Spicules are classified in the following three general classes: megascleres, microscleres, and gemmuloscleres (Ricciardi & Resiwig,

1992). Megascleres are large spicules that form the sponge’s main skeleton, while microscleres are smaller spicules (although not always present in some species) that are free-floating in the tissue. Freshwater sponges have only one type of megascleres, called oxeas, whereas microsclere spicules can vary in shape, but are typically monoaxial (Manconi & Pronzato, 2008;

Morrow & Cárdenas, 2015). Lastly, gemmuloscleres are spicules that form around the resilient coat of the gemmule (Ricciardi & Resiwig, 1992), which are asexually produced, spherical- shaped resting bodies of the sponge (Penney & Racek, 1968). These gemmuloscleres are typically required for correct species identification, as the variations in size and morphology are often unique to that species (Penney & Racek, 1968). However, gemmules are only formed at certain times of the year, hindering identification if they are not present.

1.3 Reproduction of Freshwater Sponges with a Focus on Asexual Reproduction

Freshwater sponges can undergo two types of reproduction: sexual and asexual. In

Figure 1, both reproductive cycles are shown for Ephydatia fluviatilis, a common and model species for freshwater sponges. Sponges in general, are hermaphroditic, but the timing in production of sperm and eggs is often separate, where an individual may even be one sex for a year then switch the following year (Gilbert & Simpson, 1976). Sexual reproduction occurs when male sponges release sperm into the water column and fertilize eggs in nearby female individuals, where they will begin development of larvae. Sponge larvae will leave the maternal sponge and settle on substrata before beginning to metamorphize into a fully, functioning adult

(Funayama et al., 2005). In the asexual cycle, adult sponges produce gemmules, a process triggered by environmental cues (e.g., temperature, water level) and are typically formed toward the end of the growing season and before onset of harsh environmental conditions such as cold temperatures or desiccation (Bart et al., 2019). The process of gemmule formation is usually inhibited by a factor from the parent sponge tissue. The parent sponge postpones this process until these environmental cues occur (Funayama et al., 2005). As the parent tissue begins to degrade, likely cued by abiotic environmental factors, these inhibitors, which postpone the germination (hatching) process, are eliminated (Funayama et al., 2005). Cues for germination, are similar to those that initiate gemmule formation, and include a significant change in temperature, while other factors may be included but not as heavily studied such as: light exposure, pH, water level, and water turbidity. Following these environmental cues, as well as, increases in certain metabolites (such as retinoic acid) have also been documented and may be part of a larger physiological change, but this has been investigated in only a few species

(Funayama et al., 2005; Harsha et al., 1983; Ilan et al., 1996; Loomis et al., 2008; Simpson, 1984;

Wilkinson, 1980).

Gemmules are spherical, about 100-1200 μm in diameter, and contain totipotent cells that can rapidly proliferate and are able to form a new adult sponge when conditions are suitable again (Manconi & Pronzato, 2008). This allows sponges to re-establish the population in a habitat suitable to host the mature form. The totipotent cells are able to overcome extreme environmental conditions, which would typically kill the maternal, adult sponge (Manconi &

Pronzato, 2008). Depending on the climate, in some regions, most freshwater sponges are typically found in their resting stage of gemmules with limited time represented by their adult sessile form (Manconi & Pronzato, 2007). Cells within gemmules are in a state of endogenously controlled diapause development arrest until germination is initiated (Loomis et al., 2008).

A unique feature of freshwater sponges is that they are relatively amenable to maintaining them in the laboratory. While sponges typically do not survive in the laboratory beyond two to three weeks, they can be grown from a gemmule to a fully formed, albeit small, individual sponge on a glass slide. Because of this trait, freshwater sponges have been the target of many physiological and evolutionary-development studies (e.g., Errekes-Medrano et al.,

2014). These traits also make freshwater sponges well-suited to study the relationship between the host life cycle (section 1.4 below) and the microbiome (section 1.5 below), particularly in the formation of gemmules and subsequent germination into a new sponge, the primary goal of the present study.

1.4 Sponge Cell Types, Structures, and Development from a Gemmule

There are multiple types of cells present in sponges, most of which are defined by their morphology. However, the literature has not described in detail the majority of these and mainly focuses on only a few. The major cell types of sponges are listed in Table 1 (Funayama et al., 2005; Simpson, 1984). These cell types each exhibit distinct and well-defined morphology that is fairly similar across different species, thus making them easy to identify and study

6 widely. In addition to their stable morphology, their function is consistent in the role they play in sponge development and their location within the sponge body plan. Below, I highlight major structural components of sponges and gemmules, and the relevant cell types involved. These components are likely important in both the formation of gemmules and the germination process; however, the latter process is better characterized and is the focus in this section.

Sponge gemmules are characterized by two distinct features: an acellular envelope or coat, and a mass of internal cells containing totipotent cells and other cell types, including archeocytes, that have an abundance of cytoplasmic reserve substances (Simpson & Fell, 1974).

The theca, or gemmule coat, is comprised of a spongin protein covering that protects the cells within it. There can be one, two, or three layers that create the theca and siliceous spicules may be wholly or partially embedded in it, as well. The theca layers are clearly separated and defined in freshwater gemmules, with internal, pneumatic, and external layers. The morphological differences in the layers can also help identify freshwater sponges. Some species exhibit simple structured gemmules with only one or two layers, while others exhibit a gemmule with a tri-layered pneumatic theca or one that displays spiny gemmuloscleres radially arranged around the coat (Manconi & Pronzato, 2007).

Gemmules only start to ‘hatch’ when delivered environmental ques that the adult sponge tissue could survive. The process of a gemmule hatching is called germination. There are roughly six stages of active germination in E. fluviatilis (Funayama et al., 2005) as seen in Figure

2. Stage 0 consists of the gemmule in a resting state and can be in this stage for many months, even years. Upon active germination, stage 1 typically occurs around day 2 when cells are actively migrating from the gemmule and differentiating into epithelial cells to form an envelope. Around days 3 and 4, the gemmule enters stage 2. Stage 2 consists of cells moving into the epithelial envelope, from stage 1, to proliferate and start differentiating into several different types of cells. Stage 3 of germination takes place when the sponge is about 5 days old.

During this time the aquiferous system within the sponge is formed, as well as, choanocyte chambers (chambers with flagellated collar cells for filter feeding). During this stage, a key process in the development occurs, which is the formation of the skeleton during the vegetative growth phase (Bart et al., 2019; Funayama et al., 2005) (Figure 2). Once the sponge is about a week old, stage 4 begins, where the sponge becomes fully functional with the formation of the oscula (an opening where waste water is expelled). The final stage of germination is stage 5. In stage 5 the sponge maintains its current body form and gradually increases in size from this point on (note the scale bar in Figure 2 for size of the germinated sponge).

Some of the cellular processes occurring during each stage of germination has been documented; however, there are still few studies overall on this topic with many physiological components not yet described. When germination is initiated, totipotent stem cells contained within the gemmule start to differentiate into different cells types (Höhr, 1977). These totipotent stem cells are referred to as thesocytes. Thesocytes are a double nuclei cell rich in vitelline platelets, or storage organelles, whose purpose is to provide energy for the archeocytes

(Imsiecke et al., 1995). Thesocytes start to undergo mitosis within the gemmule coat and become archeocytes (stem cells, mononucleated), which further differentiate into histoblasts

(nuclei lacking nucleolous, containing few or no vitelline platelets) (Imsiecke et al., 1995;

Simpson, 1984). These cells (histoblasts and archeocytes) start to migrate outwards from the gemmule coat onto the substrate in which the sponge is developing on. Archeocytes also differentiate into various cell types- including choanocytes and gametes (Harrison et al., 1977;

Höhr, 1977; Rozenfeld, 1970). Archeocytes are known to contain stored cytoplasmic reserve substances such as RNA, lipids, and polysaccharides that help nurture the sponge as it develops

(Ruthmann, 1965). While in the early stages of germination, the histoblasts are thought to become the basopinacocytes (basal epithelial cells) that attach the fully functional sponge to the substrate. Figure 3 represents a vertical sectioned view of a gemmule hatching sponge.

Once the sponge has become attached, the migrated cells continue to proliferate into a fully functional sponge (Funayama et al., 2005; Höhr, 1977). A key process in the development of the sponge is the construction of its skeleton, which forms the structural support for the sponge’s three-dimensional filtration system. The spicule skeleton is comprised of individual, silica spicules embedded in a fibrous organic matrix made from chitin and spongin. Spicules are produced by cells called sclerocytes. The process of skeleton formation in sponges is the only known biological mechanism in which a sequence of cooperative behaviors of individual, specialized cells lead to a self-organized biological structure using non-cellular materials

(Nakayama et al., 2010).

Sclerocytes are located at various locations within the hatching gemmule, and transported by transport cells to their final position, where they are positioned and anchored by a third type of specialized cell, called basopinacocytes (Nakayama et al., 2010). Sclerocytes have been studied in-depth and are characterized by having a well-developed rough endoplasmic reticulum, abundant mitochondria, and several small refracting inclusions

(Simpson & Vaccaro, 1974). As the cell starts the formation of a spicule, an appearance of an organic axial filament can be seen; as the process continues, the cells are flat with a large contact area to the substrate. As the spicules increase in size, they can be seen protruding from the body of the sclerocytes. Once fully formed, the spicules are embedded in a matrix of collagen-fiber bundles, often referred to as spongin (Imsiecke et al., 1995). This spongin is what helps support the surface of the sponge, which is formed by a single layer of thin, pentagonal cells called pinacocytes (pavement cells) that enclose a collagenous mesohyl (middle layer).

Once the spicules are set in an upright position within this spongin, the filter system of the sponge begins to develop, allowing the sponge to start the process of obtaining nutrients via active filter-feeding (Funayama et al., 2005). Food particles such as bacteria and single-celled algae are removed from the water and translocated into the mesohyl interior. Some of these

9 microorganisms can take resident in the extracellular region of the mesohyl, such as mobile amoeboid cells, where they become part of the commensal microbiota of the sponge host

(Gernet et al., 2005; review by Hentschel et al., 2012).

Gemmules can be hatched in the laboratory using an iterative washing protocol to isolate from the parent sponge tissue. Then stored and maintained at 4oC with no or limited light exposure for several months to a year prior to initiating active germination (see Appendix

I). In contrast to germination, there is little documentation on how the gemmules are formed within the parental sponge. Limited previous work suggests that gemmules are a pocket of resting archeocytes that have migrated within the parental sponge and formed these large masses (Funayama et al., 2005). However, how they become enwrapped by a gemmule coat is unclear. Additionally, whether symbiotic microbes are present in the gemmules, either by a selective process or as an artifact of capturing archaeocytes within the gemmules, has not been documented. Furthermore, any potential role of symbionts in the development of the sponge from gemmules has not been investigated. Rather, most research has been focused on the biology of the hatching process.

1.5 Microbiome of Phylum Porifera

Sponges, as the most basal metazoan, also offer insight into an ancient animal- microbiome relationship. Members of the phylum Porifera are hosts to diverse microbial communities with a range of symbiotic relationships including commensalism, parasitism, and mutualism (Manconi & Pronzato, 2008). Recent work has shown that sponges are selective in hosting specific lineages of bacteria, suggesting there are some selective pressures from the host in establishing and maintaining these symbiotic communities (Costa et al., 2013; Keller-

Costa et al., 2014). Furthermore, some symbionts are vertically transferred to the sponge larvae, potentially indicating evolutionary investment in maintaining certain microbial symbionts

10 across generations (Hoffmann et al., 2009; Sharp et al., 2007; Steger et al., 2008). While there are now many studies on sponge microbe phylogeny (e.g., Hentschel et al., 2002; Taylor et al., 2007;

Thomas et al. 2016), knowledge about the functions and host-interactions remains limited, particularly in freshwater sponges (Hoffmann et al., 2009).

Symbiotic microbial communities have major roles in the development, metabolism, and health of their host organisms (Lloyd-Price et al., 2016; Petersen & Osvatic, 2018). For example, in the bobtail squid, the host uses the light produced by bioluminescent Vibrio fisheri, a horizontally transmitted symbiont, to avoid predation. This squid-Vibrio system has been well characterized, where a series of developmental events is induced by the V. fisheri upon colonization in the host light organ (Reviewed by Nyholm & McFall-Ngai, 2004). The V. fisheri is able to alter the morphology of the light organ to a fully mature and functioning organ (McFall-

Ngai & Ruby, 1998). In another classic system, the pea aphid Acyrthosiphon pisum and its bacterial symbiont Buchnera, the symbionts provide essential amino acids to the host, while the host drives the production of these specific amino acids through interactions with the symbiont

(Douglas, 2011). The symbionts are intracellular, maintained in specialized cells called bacteriocytes, and are critical for the survival of the host (Gunduz & Douglas, 2009).

In sponges, some functions of the associated bacteria and archaea have been linked to defense mechanisms, vitamin production for the host, and nutrient utilization and transport, among other roles (Kiran et al., 2018). However, few of these putative roles have been empirically tested. Symbiotic Cyanobacteria, in particular, has been shown to provide photosynthetically derived carbon to the host sponge (Freeman & Easson, 2016; Wilkinson &

Fay, 1979), or in some cases, bioavailable nitrogen produced by fixation of atmospheric nitrogen

(Mohamed et al. 2010; Wilkinson & Fay 1979). Some marine sponges host Cyanobacteria within bacteriocytes similar to some insect host-microbe systems (Wilkinson & Fay 1979). Freeman and Thacker (2011) documented nutritional (photosynthate) transfer from photosymbionts,

11 namely Cyanobacteria, to the host, and found it to be highly dependent upon sponge species.

Some species obtain significant amounts of carbon and/or nitrogen from these symbionts, whereas others receive little carbon or nitrogen from symbionts (Freeman & Thacker, 2011).

Aside from photosynthetically derived carbon or nitrogen, there are other possible contributions from diverse microbes. Nitrifying microbes, which convert ammonia to nitrate as a means to produce energy, are present in many diverse sponges and may be important for removal of ammonia produced as a waste product of metabolism by the host sponge (Hoffmann et al., 2009). Nitrifying microbes and other microbes with an autotrophic metabolism, where carbon is fixed from dissolved carbon dioxide (CO2), likely support very little transfer of carbon to the host (Hoffmann et al., 2009). Additionally, marine sponges are a major source of bioactive compounds which are of interest as antitumor and anticancer compounds and these may be produced as a means of chemical defense (Blunt et al., 2004; Kiran et al., 2018; Loh & Pawlik,

2014; Mayer et al., 2010). Some studies have shown that most of these compounds are biosynthesized by sponge specific microbial symbionts (Sacristan-Soriano et al., 2011; Wilson et al., 2014). A study by Gaikwad et al. (2016) showed that bioactive compounds are also present in freshwater sponges.

Some morphological differences in cells and body structures are relevant to the ability to host microbial symbionts. For example, of the three types of gemmules produced by the freshwater species Spongilla lacustris, the gemmule with the triple-layered coat possessed very few symbionts as compared to the two single-layered coat gemmules (Simpson, 1984).

Suggesting that the additional layers provided an obstacle or deterrent for symbionts. In marine demosponges, most microorganisms were located extracellularly of the adult sponge and mostly within the sponge mesohyl. There are only a few examples of symbionts located intracellularly within vacuoles of sponge archeocytes and even within the nuclei of specific sponge cells have been observed (Hentschel, et al., 2003; Vacelet & Donadey, 1977; Wilkinson,

1978). The sponge structure with the largest effect on the microbial community is the mesohyl, which comprises all of the sponge tissue inside of the outer pinacoderm layer. The mesohyl is where most of the symbionts in marine sponges tend to reside. However, with disruption of the mesohyl, symbionts can relocate deeper in the sponge tissue (Williamson, 1979). Although, it is not known if such relocation of symbionts has any effect on the host sponge, it is interesting to note that the symbionts only seem to exhibit this shift in localization upon changes or disturbances to the mesohyl and not at random.

While there are limited studies thus far on the microbiome of freshwater sponges, these communities are likely structured differently from marine communities by (1) the location of the microbes within the sponge as freshwater sponges tend to be thin encrusting forms, (2) likely higher diversity of symbionts are present in marine sponges compared to freshwater sponges, and (3) marine microbes are highly sponge-specific and absent from most environmental libraries that have been constructed already, whereas in freshwater samples the

16S rRNA gene pool have larger similarities with its environment (Gaikwad et al., 2016; Gernert et al., 2005; Hentschel et al., 2002). The microbial community found in freshwater sponges, however, may be more specific in its localization within the sponge and so far, are only found intracellularly. For example, zoochlorellae, an alga and commonly found symbiont in freshwater sponges, occur in locations such as: (1) within cytoplasmic vacuoles in archeocytes,

(2) within choanocytes of vegetative tissue, and (3) in the thesocytes of gemmules. In a more recent study of S. lacustris, a freshwater species, the majority of its putative bacterial symbionts were found in the vacuoles of other cells, while the mesohyl appeared to be lacking any microorganisms (Gernert et al., 2005; Williamson, 1979). It is unclear whether the bacteria present in the sponge represent food particles or are symbionts and further work in localization of bacteria in freshwater sponges is necessary (Gernert et al., 2005). An interesting observation indicated that archeocytes containing a superabundance of microbes may preclude further

13 ingestion of food particles as they do not ingest as many particles as those hosting fewer symbionts. This led to the suggestion that the number of occupied vacuoles in the cytoplasm is somehow involved in the uptake rate regulation of additional particles (Van Weel, 1949). Two recent studies that leveraged high throughput sequencing, however, have identified putatively sponge-enriched and divergent lineages of bacteria in freshwater sponges (Gaikwad et al., 2016;

Gladkikh et al., 2014), lending some support to the notion of a truly symbiotic bacterial community in freshwater sponges rather than simply reflecting the same microbes as in the surrounding water.

In the present study, I conducted microbiome profiling analysis on previously collected and sequenced sponge and water samples from the New river, near Boone, N.C. (collected in

2017 and 2018). This profiling analysis leveraged well-characterized protocols for amplification, sequencing, and taxonomic analysis of the 16S rRNA gene, specific to prokaryotes

(Caporaso et al., 2018; Ul-Hasan et al., 2019;). The microbiome profiling analysis provided a preliminary overview of the composition of sponge microbiome. I then wanted to investigate a potential role of the microbiome in the development of a sponge from gemmule to adult by examining gene expression during germination with and without antibiotics. However, I found that the sponge gemmules that I collected would not hatch in the laboratory, so I shifted my project to: 1) develop a working protocol for hatching sponges in the laboratory, and 2) leverage sponges collected in the field prior to gemmule formation and post gemmule formation to examine gene expression in the microbiome during these times. To examine gene expression,

I used a metatranscriptomic approach, where total RNA from the sponge was extracted and sequenced, allowing for functional characterization based on mRNA produced by the sponge and microbes (Anwar et al., 2019; Fiore et al., 2015; Leimena, et al., 2013; Westreich et al., 2018).

Due to high natural variability in the presence of sponges and the sponge life cycle, sponges collected in 2018 and 2019, that together span the time before and after the onset of asexual

14 reproduction, were used in the metatranscriptomics analysis. This work presents the first metatranscriptomics analysis of freshwater sponges and lays the groundwork for future experiments aimed at understanding the potential role of the microbiome in the sponge life cycle.

2. Methods

2.1 Field Methods

Freshwater sponge samples were collected for three consecutive years from New river state park in Ashe County, North Carolina (36.4615o N, 81.3387o W). Sponges collected in 2017 were collected for 16S rRNA microbiome profiles and preserved in DNA buffer, whereas sponges collected during 2018-2019 were preserved in RNAlater (Invitrogen, USA) for metatranscriptome analysis. The experimental design was to collect sponges prior to gemmule formation and post gemmule formation for metatranscriptome analysis. In 2018, two individuals were collected in July prior to gemmule formation and two in August post gemmule formation. In 2019, another three individuals were collected in July and August prior to gemmule formation and one individual in September post gemmule formation (Table 2). The individuals collected prior to gemmule formation and after gemmule formation are not the same individuals.

Each sample for microbiome profiling and for metatranscriptome analysis was collected using a sterile razor blade to scrape a portion of the sponge off the rock, or substrate, into a sterile microcentrifuge tube filled with 2mL of DNAbuffer or RNAlater. Those samples collected in 2017 for microbiome profiles of 16S rRNA were stored in 2mL of DNAbuffer (Seutin et al.,

1991). Each tube was individually labeled with the date and location of each sample and placed on ice until they were returned to Appalachian State University. Samples were stored at -20oC, until further analysis. From these freshwater sponge samples one set (n=5 individuals) from pre-gemmule formation (July 2018 and 2019, August and September 2019) and one set (n=3 individuals) from post-gemmule formation (August 2018, September 2019) were used to generate the metatranscriptome dataset (Table. 2).

Environmental data including three independent water samples (500 mL each) for nutrient analysis and microbiome analysis were collected at each time point. The water samples

16 were vacuum filtered using a sterile 0.2 µm polyethersulfone filter; the filters were stored in

RNAlater and frozen, while approx. 30 mL of the filtrate was stored in a 50 mL conical tube and frozen for inorganic nutrient analysis. Inorganic nutrient analysis of anions (nitrate, phosphate, sulfate, chloride), was performed using an ion chromatograph with quantification based on an external standard curve. The analysis was performed previously by Allison Griggs with Dr.

Carol Babyak in Chemistry and Fermentation Sciences at ASU. During 2019, a subset of water samples (1mL each sample) were fixed in filtered paraformaldehyde (0.5% final concentration) and frozen for flow cytometry analysis at Bigelow Facility for Aquatic Cytometry (ME, USA).

Additionally, in 2019, temperature, pH, conductivity data were collected using a handheld multiparameter meter (Oakton, IL, USA).

2.2 Laboratory Methods

2.2.1 Preliminary Microbiome Profiling of 16S rRNA Genes

DNA was extracted from sponge and filters collected in 2017 using hexadecyltrimethylammomium bromide (CTAB) extraction. The 16S rRNA genes were amplified using the earth microbiome project (EMP) standard primer set (515F and 806R) and

EMP protocol (Caporaso et al., 2018; Ul-Hasan et al., 2019). Paired-end library construction and high throughput sequencing using an Illumina MiSeq was performed by Dr. Cole Easson at Nova

Southeastern University in January 2018. Sponge and water samples from the New river were analyzed for microbiome taxonomy using the Dada2 pipeline in R (Callahan et al., 2016), which produces amplicon sequence variants (ASVs) that differ from each other by a minimum of one nucleotide. The ASVs as well as read abundance within each ASVs were used in alpha and beta diversity metrics using Phyloseq (McMurdie & Holmes, 2013) in the R environment. Analysis of similarity (ANOSIM) was used to compare microbiome composition across samples. Further analysis and tests, such as PERMANOVA, will be performed prior to manuscript publication.

2.2.2 RNA and DNA Extractions and Sequencing for Metatranscriptome

Extraction of total RNA and DNA from each of the freshwater sponge samples was performed using a TRIzol Reagent (Sigma Aldrich, USA) per the manufacturer’s protocol for simultaneous extraction of RNA and DNA, where RNA- and DNA-containing layers are separated by centrifugation and extracted separately. The extracted RNA was then incubated with DNase I

Treatment (Promega, WI, USA) to remove genomic DNA and Monarch RNA Cleanup Kit to remove protein and other contaminants (New England Biolabs, MA, USA). DNA extraction was performed using the TRIzol Reagent standard protocol. The concentration and quality of both the RNA and DNA were analyzed via NanoDrop 2000. Due to low concentration and quality, the extracted RNA was further purified with a phenol-chloroform extraction. The acceptance of these RNA products was selected upon quality ratios of 260/280 with values upwards of 2.0 or greater, except samples 16 and 17 which had values of 1.60 and 1.35 respectively. The extracted

RNA was processed by GeneWiz (NJ, USA) to analyze quality and concentration by a bioanalyzer, remove eukaryotic ribosomal RNA (rRNA), which generally comprises the majority of RNA in a total RNA pool (Ribo-Zero rRNA removal kit (Illumina, CA, USA)), construct the paired-end cDNA library, and sequence on one lane of an Illumina HiSeq 2000 (150 nt read length).

All extracted DNA products were low in concentration (average 69 ng/µL) and were therefore selected based on the absorbance ratio of 260/280 value of typically greater than

1.40. The extracted DNA samples were processed for 16S rRNA gene amplification with the

Earth Microbiome primers (515F and 806R (Caporaso et al., 2018; Ul-Hasan et al., 2019)) and sequenced using an Illumina MiSeq.

2.2.3 Sequence Assembly of Symbiont Metatranscriptome

Symbiont sequence analysis was performed using the SAMSA pipeline (Simple Analysis of Metatranscriptomes through Sequence Annotation), version 2.0 (Westreich et al., 2018) on a high-performance cluster computer. The pipeline consists of three main processing steps

(preprocessing, annotation, and aggregation/downstream processing) and uses a set of python scripts that brings together multiple programs involved in processing, a flow chart of this process can be seen in Figure 4. For SAMSA2 to provide the best results, metatranscriptomes need to be ribodepleted (removal of ribosomal RNA) and sequenced in a paired-end format, both of which were completed with the sponge metatranscriptome dataset (Westreich et al.,

2018). The sequencing of RNA (RNAseq) produced two FastQ files for each sample (n = 8 total) were produced, containing the forward and reverse reads respectively. Here I focus on the symbiont (i.e., bacterial) metatranscriptome, specifically using the bacterial RefSeq database for annotation within the SAMSA2 pipeline, and future work will include analysis of the host sponge transcriptome.

2.2.3.1 Preprocessing

RNAseq produced forward and reverse reads for each RNA (cDNA) that was sequenced, therefore the first step in preprocessing was to merge paired reads. PEAR 0.9.8 was used to merge paired reads (Zhang et al., 2014), which produced a single file of extended fragments from the two (forward and reverse) FASTQ sequencing files per sample. The command for

PEAR was:

$pear_location/pear -f $forward_reads.fastq -r $reverse_reads.fastq -o $out_name

Where, -f points to the forward reads, -r points to the reverse reads, and -o designates the output file of merged reads. Next, low-quality sequences and/or adaptor contamination were removed using the program Trimmomatic 0.36 (Bolger et al., 2014). Trimmomatic was specifically designed to be used for Illumina sequences and thus includes the adapter and primer sequences for these machines. Trimmomatic was also used to remove low quality reads

(below phred = 33) or trim the reads to produce an average phred score of 33. The command for Trimmomatic was as follows:

java -jar $./trimmomatic-0.36.jar SE -phred33 $infile $outfile_name SLIDINGWINDOW:4:15

MINLEN:99

Where, SE is the flag for single end mode (only one input file-the merged reads file from

PEAR), -phred33- specifies the base quality encoding (FASTQ quality score), SLIDINGWINDOW-

:, and MINLEN- specifies the minimum length for reads to be kept. For the sliding window, the ‘windowSize’ specifies the number of bases to produce an average as the program moves across the read, and ‘requiredQuality’ specifies the average quality threshold. The phred score is the quality score of base quality in DNA sequencing.

A ‘cleaned’ output file for each sample is produced from Trimmomatic for the final preprocessing step. Although a prior ribodepletion has already been performed, due to the naturally high rRNA counts present in microbiome samples, a digital ribodepletion is included in the SAMSA2 pipeline as the final preprocessing step. SortMeRNA 2.1 (Kopylova et al., 2012) removes any ribosomal reads left in the samples that may affect annotation of the metatranscriptome. SortMeRNA incorporates several reference databases for both bacterial and

20 eukaryotic sequences for rRNA identification: SILVA, GreenGenes, RDP. The command for

SortMeRNA was as follows:

$sortmerna_location/sortmerna --ref $sortmerna_location/rRNA_databases/silva-bac-16s- id90.fasta,$sortmerna_location/index/silva-bac-16s-db --reads $file --aligned $file.ribosomes -- other $file.ribodepleted --fastx --num_alignments 0 --log -v

Where, --ref points to the FASTA reference file, --reads points to the FASTA/FASTQ reads files, --aligned points to the aligned reads filepath, --other points to the rejected reads filepath, --fastx indicates to make the output file in a FASTA/FASTQ format, --num_alignments 0 indicates all alignments reaching the E-value threshold are reported; very slow, --log creates an output of overall statistics, and -v stands for verbose. Two files were produced: $file.ribosomes which contains all the rRNA sequences, and $file.ribodepleted which contain all the reads not identified as ribosomes. The ribodepleted file was used in the annotation step.

2.2.3.2 Annotation

SAMSA2 uses a BLAST-like aligner called DIAMOND (Buchfink et al., 2015) for the annotation process. This is because DIAMOND was designed to handle the large numbers of reads metatranscriptomes may possess at a rate 10,000 times faster than BLASTX (Westreich et al., 2018). DIAMOND has the ability to annotate sequences against one or more reference databases. The databases selected were the NCBI bacterial RefSeq (Tatusova et al., 2014) and

SEED Subsystems hierarchical databases (Overbeek et al., 2014) for functional activities. The commands for DIAMOND were as follows: to perform the annotation process and to covert the generated results file into a data table for the aggregation step, respectively.

(1) $diamond_location blastx --db $diamond_database -q $file -a $file.RefSeq -t ./ -k 1 –

sensitive

(2) $diamond_location view --daa $file.RefSeq.daa -o $shortname -f tab

Where, in command 1 , -db specifies the reference database to use, -q indicates the query file name, -a indicates the name of results file in DIAMOND format, -t sets the temporary directory locations needed as the program runs, -k 1 indicates the number of hits above cutoff

(e-value cutoff = 0.001) to return, and settings changed from “–fast” to “–sensitive” to increase accuracy. In command 2, --daa points to the output file in .daa format, -o determines the output file name, and -f indicates separator of values in final output.

2.2.3.3 Aggregation and Downstream Processing

DIAMOND produced annoated files are returned in line-by-line results, each sequence from the metatranscriptome that had a corresponding match with a reference base sequence occupied one line in the output file. The next step was to use Python 2.7 scripts to aggregate these line-by-line files into condensed, summary tables (Westreich et al., 2018). Python is able to create two files from the same DIAMOND output reads: one for organism annotations and the other for functional annotations into 3 column tables. The 3 columns are broken down into 1) percentage of total reads, 2) read count, 3) annotated organism or function. This is generated by the following two commands:

(1) python $python_programs/standardized_DIAMOND_analysis_counter.py -I $file -D

$RefSeq_db -O

(2) python $python_programs/standardized_DIAMOND_analysis_counter.py -I $file -D

$RefSeq_db -F

Where, -I points to the infile, -D indicates the specific reference database to search against, -O (in command 1) indicates to aggregate all reads by organism, and -F (in command 2) indicates to aggregate all reads by function. In addition, Python can also produce hierarchical annotations using SEED Subsystems to include abundance counts with all hierarchy level information, using the command:

python $python_programs/DIAMOND_subsystems_analysis_counter.py -I $file -D $Subsys_db -O

$file.hierarchy -P $file.receipt

These count files are then saved as the final product of the automated workflow for the

SAMSA2 pipeline before moving into R for the statistical analysis portion (Westreich et al.,

2018).

2.2.3.4 Additional Applications

SAMSA2 also has the ability to annotate functions by specific organisms, using a Python program named “DIAMOND_specific_organism_retreiver.py”. This is because each processed read already receives both an organism and functional annotation. The command for the specific organism retriever was:

python $python_programs/DIAMOND_specific_organism_retriever.py -I $file -SO (genus or species) -D $RefSeq_db

Where, -I points to the input file, which is the DIAMOND results file from annotation step above, -SO specifics which specific organism, either genus or species, to be isolated, and -D specifies the reference database to use. This creates an output file containing all the individual transcripts that originated from the specified organism. Following this, the produced file will have to be aggregated as well using the command:

python $python_programs/DIAMOND_analysis_counter.py -I $file -D $RefSeq_db -F

Where -I points to the infile, -D indicates the specific reference database to search against, and -F indicates to aggregate all reads by function. These files were saved as the final product and ready for further statistical analysis using R (Westreich et al., 2018).

2.2.3.5 Statistical Analysis and Visualization of RNAseq

All statistical analyses were performed in RStudio version 3.6.3 using R scrips from

SAMSA2 pipeline. These scripts use were used to calculate Shannon and Simpson diversity indices, taxonomic and functional bar plots, and principle coordinates analysis (PCA) analysis.

The R scripts also included a specific set of scripts that utilize the program DESeq2 to calculate differential features (i.e., taxa or functions) between pre and post gemmule samples. The pre gemmule samples are: samples 16, 17, 50, 55, and 58; whereas, the post gemmule formation samples are 24, 25, and 57. DESeq2 is a differential expression analysis which takes read count data from these samples and performs an internal normalization, calculating a geometric mean for each gene across all samples, and uses a Wald test to calculate statistical significance and

Benjamini and Hochberg test correction for multiple testing (Love et al., 2014).

The resulting profile of organisms in each sponge sample, as well as, the functional profiles of gene expression in the sponge host and microbiome were first characterized as an

24 overview of “who is there” (i.e., taxonomic profile based on RefSeq database taxonomy) and

“what are they doing” (i.e., functional profile of transcripts) prior to, and during the time in which gemmules developed in the sponges (“Pre” and “Post” gemmule formation).

Differentially abundant features were calculated for taxa and functional annotation using RefSeq and functional annotation using the SEED Subsystems (Overbeek et al., 2014). A subsystem in the SEED database is a collection of functionally related proteins and the SEED subsystem organization provides hierarchical functional categories, facilitating comparisons in function between groups. Specific taxa that were identified as differentially abundant between the two time points were used in organism-specific analysis described above.

2.2.4 Environmental Data

In 2017 and 2018 various environmental data was collected for New river and Jacob

Fork river including: phytoplankton and bacteria cell densities, pH, temperature (oC), conductivity, and chloride, nitrate and sulfate concentrations. These datasets were small with n=3 samples each. Therefore, a Wilcoxon test was performed in R for all environmental comparisons (e.g., temperature or pH between sites, phytoplankton abundance vs. bacterial abundance at a given site).

3. Results

3.1 Sponge Identification and Preliminary Microbiome Profiling of 16S rRNA Genes

Morphological and spicule preparations of the collected sponges indicate a putative identification of the sponges as Ephydatia fluviatilis with the megascleres being slightly curved, and mostly smooth oxea. Gemmuloscleres were unable to be visualized; therefore, genetic identification is in the process of being performed.

Total input reads ranged from 38,340 to 95,249 with an average of 64,856 across all 27 samples sent for 16S microbiome profiling (Table 3). Between input reads and nonchimeric reads, the percentage loss was an average of <27% for all samples. With the lowest number of nonchimeric reads being 7,813 but the average for all being 17,573.

Microbial community profiling using 16S rRNA genes indicated overlap in taxonomy at the phylum level between water and sponges with the major taxonomic groups including

Proteobacteria, Bacteroidetes, and Actinobacteria (Figure 5). However, at low taxonomic resolution, analysis of similarity (ANOSIM) applied to the sponge and water ASVs communities indicated a significant difference in composition the level of operational taxonomic units (ASVs, as with operational taxonomic units (OTUs) these are a taxonomic unit loosely used as a bacterial ‘species’) between the two habitats (ANOSIM, R=0.23, p=0.01). Time of sampling (July,

August, September) was also shown to have an effect on community structure in all samples

(ANOSIM, R=0.54, p=0.001), which is clearly visualized by network analysis showing the separation of water and sponge communities with those being more similar indicated with shorted and thicker lines. As well as, samples in July, both wate and sponge, are completely isolated in comparison to August and September (Figure 6).

Taxonomic comparisons at the ASV-level, however, highlighted certain taxa that were detected only in the sponge or water (Figure 7). There was a total of 594 ASVs produced, with

312 ASVs in the sponges and 415 ASVs detected in the water. Of the ASVs observed in the sponges, 189 (60%) were present only in the sponges and not in the surrounding water. The

ASV table was not rarefied (McMurdie & Homes, 2014). Specifically, within the most abundant

300 operational taxonomic units (ASVs) at the phylum level, several ASVs were observed to be abundant in the sponge samples while being absent from the water samples (Figure 7), thereby identifying theses taxa as putatively enriched or specific to sponges. Some of the ASVs specific to sponges included taxa within Classes Flavobacteriales and Sphingobacteriales within the phylum Bacteroidetes (Figure 8), as well as Betaproteobacteriales, Xanthomonodales, and

Sphingomonodales within the phylum Proteobacteria.

3.2 Metatranscriptome Results

Total reads generated per samples ranged from 54,321,279 to 37,446,129 (Table 4).

Between raw reads and final read counts, the percentage lost was <31% for all samples with an average loss of 25% after the merging of paired reads and removal of low-quality reads. With the lowest number of final reads over 27,000,000 and the most being over 40,000,000 (Table

4).

In 2018, a total of four individuals were collected between July and August. The two collected in July showed no evidence of gemmule formation, whereas the other two collected in

August gemmules were present. In 2019, another four individuals were collected over the course of three months (July, August, and September) (Table 2). The resulting profile of organisms in each sponge sample, as well as, the functional profiles of gene expression in the sponge host and microbiome were first characterized as an overview of “organism” (i.e., taxonomic profile based on RefSeq database taxonomy) and “function” (i.e., functional profile of transcripts based on RefSeq database gene annotations) prior to, and during the time in which gemmules developed in the sponges (“Pre” and “Post” gemmule formation). Shannon and

Simpson diversity metrics were similar between the pre-gemmule (Pre) and post-gemmule

(Post) sponge samples. For organismal results the Shannon diversity for Pre was 6.12 and for

Post it was 6.18, and the Simpson diversity was 0.98 for Pre and 0.98 for Post. For functional results, Pre was 4.65 for Shannon and 0.87 for Simpson and Post was 4.67 for Shannon and 0.88 for Simpson (Table 5).

As an overview, PCA plots, which visualize the community composition data (taxonomy or function profile and abundance) in a two-dimensional space, did not show separate groups for Pre and Post sponge samples by organism taxonomy (Figure 9) or by functional annotation from the RefSeq database (Figure 10). The lack of separation of the Pre and Post sponge samples is supported by the results of a permutational multivariate analysis of variance test between the two groups (ADONIS: R2=0.069, p=0.84 by organism; R2=-0.068, p=0.74 by function). In both analyses, organism and function, although particularly for function, the Pre samples tended to be more dispersed than the Post samples (Figures 9, 10). However, there are also more Post samples than Pre samples.

The most prevalent taxa across all sponge samples based on the metatranscriptome analysis included groups mainly within the Bacteroidetes and the Class Gammaproteobacteria

(Phylum Proteobacteria). Differential abundance of taxa from Pre to Post samples was detected with DESeq2 analysis (Table 6). The information from DESeq2 includes the means of both sample types (Pre and Post), log2FoldChange, lfcSE (error associated with the fold change), the statistic, p-value, and adjusted p-value. The log2FoldChange is the indicator as to whether the organism increased or decreased in abundance. A negative log2FoldChange means there was a decrease in abundance from Pre to Post. A positive log2FoldChange means there was an increase. Organisms that showed differential abundance between time points include

Actinobacteria including taxa within the Family Streptomycetaceae (i.e., Kitasatospora and

Streptomyces genera), Flavobacterium within the Bacteroidetes phylum, Betaproteobacteria

28 such as Polynucleobacter spp., Alphaproteobacteria such as Sphingomonas spp., an

Epsilonproteobacteria, Sulfurimonas, and one taxon within Firmicutes (Bacillus). There was no clear trend for a broad taxonomic group (e.g., all Flavobacterium and all Polynucleobacter) to increase and decrease (respectively) in abundance from Pre to Post gemmule time points.

The most abundant functions detected in the sponge samples, based on RefSeq functional annotation were involved in energy and maintenance metabolism such as electron transport chain enzymes and the chaperone protein DnaK (Figure 11). Note that proteins of unknown function were also prevalent (Figure 11). Differential abundance of specific functional annotations was detected for 12 functions assigned from RefSeq between Pre and Post samples, where five functions decreased and seven increased (Table 7). Functions that increased in abundance in Post relative to Pre samples include transcripts for catechol 2,3-dioxygenase, involved in aromatic compound degradation, porin proteins, which are channel proteins specific to certain molecules and common in Gram negative bacteria, disulfide bond formation protein B, which catalyzes disulfide bond formation in periplasmic proteins (thus would be found in Gram negative bacteria), enzymes in nitrogen metabolism (type III glutamate— ammonia lyase), pilus assembly protein, a domain of unknown function (DUF4293), and an elongation factor involved in proteins translation and ribosome recycling (EF-G) (Table 7). In contrast, functions that decreased in abundance in the Post samples included transcripts involved in vitamin B12 bioynthesis (Uroporphyn-III methyltransferase), amino acid metabolism (bifunctional o-acetylhomoserine/o-acetylserine sulfhydrylase), as well as different porin proteins, the chaperone protein Hsp70, and a hydrogenase protein (Table 7).

Some similar functional profiling results were observed with the SEED subsystem categories although at a greater scale (Figure 12). Looking at Figure 11, the top bar graph shows the relative activity of total sample, which is predominantly ‘hypothetical protein’ and

‘other’, as well as a couple subunits of cytochrome c oxidase (I, II, and 3) and NADH-quinone

29 oxidoreductase subunit H, and these appear relatively consistent across both Pre and Post.

Differential abundance of SEED subsystems indicated a marginal but statistically significant increase in respiration, cofactors, virulence, and photosynthesis and iron acquisition and metabolism (Table 8). SEED subsystems that decreased in abundance in Post samples included motility and chemotaxis, cell wall and capsule, transposable elements, sulfur metabolism, and dormancy and sporulation (Table 8).

Lastly, from the list of taxa that showed a significant difference in abundance between

Pre and Post gemmule samples (Table 6), four genera were selected for more in-depth functional analysis: Polynucleobacter, Flavobacterium, Sulfurimonas, and Sphingomonas, as they were the most prominently listed or were of interest for their potential role in nutrient cycling.

They were also of interest due to also being prominent in the 16S data. These four were analyzed for the functional annotation of transcripts within just these organisms using the NCBI

RefSeq database. However, within these four organisms, only two genera displayed functions that were significantly different from pre and post samples. Those two genera were

Polynucleobacter and Flavobacterium (Table 9). There were seven functions in Polynucleobacter that were different with six of them increased in abundance from Pre to Post gemmule formation, which were all involved in bacteriochlorphyll biosynthesis or one enzyme, transaldolase, involved in the metabolism of carbohydrates in the pentose phosphate pathway

(Table 9). In contrast, Flavobacterium had 25 functions that all of them significantly decreased in abundance, including proteins likely involved in maintenance metabolism (e.g., acyl-coA ligase in fatty acid metabolism, NADH-quinone oxidoreductase subunit N) as well as several proteins of unknown function and proteins known to be involved in protein-protein interactions in a host (i.e., fibronectin type III domain-containing protein, ankyrin repeat domain-containing protein, Type IX secretion system (T9SS C-terminal target domain- containing protein)) (Table 9).

3.3 Environmental Results

Flow cytometry of water samples yielded an overview of phytoplankton and bacterioplankton abundances in the water, both of which may be food for the sponges. A comparison of two locations where sponges have been collected, New river and the Jacob Fork river which is at a lower elevation near Hickory, NC, are provided here to add context for the environmental data. Phytoplankton density was always lower than bacterial density by about an order of magnitude (Figure 13). Phytoplankton were more variable in abundance at the New river, with a sharp increase in August (Figure 13). There was no difference in phytoplankton abundance between the New river and Jacob Fork river (W= 57, p-value=0.1615). Bacteria was significantly more abundant in New river than Jacob Fork (W=0, p-value= 4.1145e-05).

Temperature, pH, and conductivity were not significantly different between the two locations (Figure 14a,b,c). Inorganic nutrients were generally low across time points and between locations (Figure 15). Nutrient analysis for 2019 has not yet been conducted, but data from 2017 and 2018 indicate low anion nutrients in those two years (Figure 15).

4. Discussion

Microbial symbionts of metazoans can have a significant role in the development of host structure, immunity, diet and ecological niche (Brestoff & Artis, 2013; Douglas, 2011; Hentschel et al., 2012; Manconi & Pronzato, 2008; Reviewed by Nyholm & McFall-Ngai, 2004). In sponges most microbial symbionts are considered to be commensals, although experimental work has mainly been focused on photosymbionts in sponges, for which it is fairly straightforward to identify potential host benefits. Here, I set out to investigate how the composition of these microbial communities change as the sponge undergoes developmental changes involved in asexual reproduction, and determine what role, if any, the microbiome may have in the facilitating developmental change of the sponge host. An experimental manipulation of the sponge microbiome was the original goal; however, due to experimental challenges, I shifted my work to a correlative study and acknowledge both the limitations in interpretation of my results and the foundation that this work provides for future experimental studies.

4.1 16S rRNA Gene Microbiome Profiling

Strong evidence was found in the 16S rRNA gene profiling results of this study that the microbial communities in sponge and water are significantly different from each other.

Specifically, while some groups overlap between sponge and river water, there were groups of taxa were that were present only in the sponges as well as taxa that were abundant in the sponge and rare in the water. This distinction of the sponge microbiome from that of the surrounding water is similar to findings from other studies on marine and, more recently, on freshwater sponges (Gaikwad et al., 2016; Reveillaud et al., 2014). Furthermore, of the taxa observed in sponges in the present study, the best matches to these ASVs in the NCBI database were to taxa known to be symbionts in other sponges. Specifically, this includes ASVs within the phylum Bacteroidetes (Orders: Chitinophagales, Cytophagales, Flavobacteriales, and

Sphinobacteriales) and Proteobacteria (Classes: Gamma- and Alphaproteobacteria). Other taxa found to be more prevalent in the sponge host compared to the water were similar to those in other freshwater sponge studies that used high throughput sequencing methods:

Actinobacteria, Bacteroidetes, Chloroflexi, Cyanobacteria, Nitrospira, and Proteobacteria (Alpha and Gammaproteobacteria) (Gaikwad et al., 2016; Gladkikh et al., 2014). A notable exception is the low abundance of Cyanobacteria that was observed in the sponge and water in the present study, similar to several other freshwater sponge studies (Costa et al., 2012; Gladkih et al., 2014), but in contrast to observations by Gaikwad et al. (2016). The compositional results here are also similar to a study performed on the 16S rRNA microbial profile of the cosmopolitan freshwater sponge Spongilla lacustris whose gene library comprised four major lineages:

Alphaproteobacteria, Actinobacteria, Betaproteobacteria, and Chloroflexi (Gernert et al., 2005).

These data overall support both high diversity and potentially divergent symbiotic bacterial taxa within freshwater sponges.

4.2 Metatranscriptome Analysis of Sponge Symbiont Communities

In the metatranscriptome analysis, SEED produced a pie chart showing the large scale of the top 10 functional subsystems functions that were prevalent. The biggest portion of that pie chart is indicated by these three functions: protein biosynthesis, electron accepting reactions and electron donating reactions (Figure 12). RefSeq functions bar chart are more specific gene functions rather than broad scale. The majority of these include hypothetical protein (gene of unknown function), as well as, energy functions being a major component (cytochrome c and

NADH-quinone oxidoreductase part of electron transport chain) (Figure 11). Molecular chaperone DnaK is also a key function to note, this falls under housekeeping protein folding in addition to being involved in DNA replication and energy & maintenance processing.

Pre and Post microbial communities did not significantly separate into two different groups based on taxonomy (organismal analysis), but we did see specific taxa increased in one or the other group. Some taxa that were identified as significantly different in abundance between the Pre and Post samples, overlapped with taxa identified as enriched in the sponges based on 16S rRNA gene profiling. These include taxa within the Orders Flavobacteriales and

Sphingomonadales, both of which increased in abundance from Pre to Post metatranscritptome samples and were taxa observed only in the sponges based on 16S microbiome gene profiling

(Figure 8). Separately, two taxa that were more abundant in Pre samples include the genera

Streptomyces sp. and Xenorhabdus sp. These are of interest as Streptomyces is a recognized symbiont in multiple organisms, including marine sponges (reviewed by Seipke et al., 2012).

This group is of particular interest because of the bioactive molecules and antimicrobial properties they may produce (Almeida et al., 2019; Reviewed by Seipke et al., 2012).

Xenorhabdus sp. also produces bioactive compounds with a broad-spectrum of antimicrobial activity and are well known for their role as symbionts of entomopathogenic nematodes

(Dreyer et al., 2018). Thus, these taxa in particular, (Flavobacteriales, Sphingomonadales,

Streptomyces, Xenorhabdus) may represent symbiotic (facultative or obligate) bacteria within the sponges.

Another bacterial group that was prevalent in the results of the differential abundance analysis, was that of the genus Polynucleobacter, all of which decreased in abundance from Pre to Post samples. Polynucleobacter was not highlighted as a taxon of interest in the 16S microbiome samples, which is interesting to note. Although, that doesn’t mean they weren’t present and further investigation is needed to explore that idea. Polynucleobacter was originally described as endosymbionts that resided in ciliates (P. necessaries, Hechmann & Schmidt, 1987), but more recently, multiple Polynucleobacter strains were identified as abundant free-living

34 planktonic bacteria in freshwater systems (Hoetzinger et al., 2018). I cannot determine from these data alone whether the strains observed here are free living or symbiotic.

4.3 A Potential Framework for Symbiont-Host Interaction in Freshwater Sponges

While it is not possible at this point in time to determine the nature of the relationship for the major taxa observed in the metatranscriptome analysis, I have organized the results into a framework to generate hypotheses for future work (e.g., tracking microbiome changes over time, experimental manipulation of the microbiome in the laboratory). There are three main categories of bacterial life histories that could be present in the sponge samples: 1) free-living,

2) obligate or long-term facultative symbiont that is always present in the sponge and likely to be present in the gemmules and gametes, and 3) short-term facultative symbiont that disperses from the sponges when it disappears.

First, if we are detecting free living bacteria that happened to be included in the sample from water present in the sponge or as food particles, we might expect the following: They would likely exhibit maintenance metabolism that may not change over the two time points sampled here and living in the water, they are likely to be photoautotrophic or heterotrophic. In these bacteria, we may not see changes in metabolic activity over two time points other than photosynthesis genes that may change in expression as irradiance and temperature changes between July and September. In comparison to symbiotic bacteria, there would likely be fewer, or lack of, motility and chemotaxis transcripts, as well as transcripts for transposable elements, viral defense, and proteins with eukaryotic-like domains (Thomas et al., 2016).

The results presented here within the functional analysis of the Polynucleobacter genus indicated the only significant changes in transcript abundance were those associated with photosynthesis, which increased in the Post samples. This could be driven by environmental factors (e.g., irradiance, temperature) and may indicate some or all of these taxa are free-living.

No other symbiotic traits (e.g., eukaryotic-like domains, chemotaxis) were observed in the

Polynucleobacter analysis. However, this does not preclude some of these strains from being symbiotic; they may be at low abundance or may not change their activity with changes in the sponge, and several of the best matches in NCBI were to the symbiotic P. necessaries specifically.

Further monitoring of sponge and water microbial populations should shed more light on the role of this group.

Second, in contrast to free living bacteria, if we are detecting facultative or obligate symbiotic bacteria that associate with the sponge during vegetative growth and are intentionally transmitted into the gemmules, we might expect that some of the bacteria that would be transmitted into the gemmules are also still present in the remaining sponge tissue.

Such a relationship would be similar to vertical transmission in many marine sponges which can include high percentages (>30%) of symbionts that are passed to the offspring through gametes (Reveillaud et al., 2014; Schmitt et al., 2012; Webster et al., 2010). In this case, we would expect a certain set of traits to be present and potentially change in expression with the developmental and structural changes in the sponge: 1) bacterial metabolisms that include photoautotrophic, and potentially ammonia oxidizing bacteria and sulfur oxidizers (using waste products from the sponge), as well as heterotrophic bacteria (Liu et al., 2011; Moitinho-Silva et al., 2017; Slaby et al. 2017). If these microbes have a long-term and intimate association with the sponge, then the metabolic activity may show no change over time or may increase as sponge begins to disappear and undergoes structural changes. An increase in metabolic activity may be the result of more nutrients released as the sponge degrades or as a result of interaction with and movement within the sponge. 2) Similarly, metabolic functions that could provide certain nutrients (e.g., vitamins, amino acids, organosulfur compounds) to the host may show no change or an increase in expression as the host produces gemmules and begins to disappear.

3) Presence of transposable elements and viral defense would be expected to be higher in the

36 symbiotic community than in free living microbes based on marine sponge microbiome analyses (Horn et al., 2016; Thomas et al., 2016), but may show little to no change over the time points. 4) Disappearance of the sponge tissue may also increase expression of motility and/or chemotaxis genes in symbiotic bacteria as these bacteria respond to remodeling and developmental changes in the sponge. 5) Expression of proteins with eukaryotic-like domains has been documented as an enriched trait in sponge and other symbiont communities in comparison to free living microbes (Horn et al., 2016; Slaby et al., 2017; Thomas et al., 2016).

Expression of such proteins might decrease upon sponge degradation after gemmules are produced if these bacteria leave the sponge (i.e., facultative symbionts) or may increase if this is involved in transmission into gemmules (i.e., obligate, or at least long-term, symbionts). 6)

Lastly, expression of nutrient transporters might increase in expression as the sponge begins to disappear due to longer commitment to living in a sponge and potential release of metabolites as the sponge degrades.

Relevant results that may apply under this second framework would be functions highlighted in Table 7, such as: pilus assembly protein, porin family protein, catechol 2,3- dioxygenase, and type III glutamate-ammonia ligase. These functions increased in expression in the Post samples. Pili are structures used to adhere to surfaces, like a host, or in bacteria- bacteria-bacteria communication and have been documented in marine sponge symbiotic bacteria (Liu et al., 2011). Therefore, an increase in pilus assembly protein could be evidence of increased bacteria-bacteria interaction or potentially bacteria-host interaction in the Post samples. The Porin family of proteins is a type of transport protein usually located in the outer membranes of Gram-negative bacteria, which were the dominant type of bacteria in the New river sponges. The main function of porins is to import specific molecules into the cell, such as certain types of sugars as well as secondary metabolites, any of which would likely to be produced within the sponge (Fan et al., 2012; Kamke et al., 2013). In particular, as the sponge

37 degrades, it may release different metabolites and nutrients, and these bacteria may be well- suited to take advantage of such compounds. Similarly, the increase in transcripts for catechol

2,3-dioxygenase in the Post samples may make sense as this enzyme is involved in degradation of aromatic compounds that could be released by the sponge and/or other bacterial symbionts.

Type III glutamate-ammonia ligase, also known as glutamine synthetase, is an important enzyme in nitrogen regulation which forms glutamine from ammonia and glutamate. Nitrogen regulation through the protein PsII, which regulates glutamine synthetase has been documented as enriched in marine sponge symbiont communities in comparison to free-living communities (Moithinho-Silva et al., 2017; Thomas et al., 2016). While it is still speculation, the observed increase in transcript abundance of these functional traits in the Post samples may reflect characteristics of a group of bacteria that are committed to residing in the sponge and have adapted to their host environment and life cycle. Potentially these bacteria would be candidates to be transferred into gemmules, if there is selective pressure to maintain these lineages with the sponge life cycle.

Third, if we are detecting facultative symbiotic bacteria that associate with the sponge during vegetative growth and then transition to free-living after the sponge disappears, we might expect that these bacteria would either leave the sponge to become free-living again or remain until they are released as the host disappears. In this case we would expect similar traits to the long-term symbionts described above, but with some potential differences in response to gemmule formation and degradation of the sponge: 1) Such symbionts could have similar metabolic potential as the long-term symbionts described above (i.e, photoautotrophs, potentially ammonia oxidizing bacteria and sulfur oxidizers, and heterotrophs). However, there may be more varied responses to sponge degradation. Heterotrophs may increase in metabolic activity as sponge disappears, while ammonia and sulfur oxidizing may decrease if reliant on host metabolism or increase as sponge tissue degrades. 2) The expression of motility and/or

38 chemotaxis genes might increase in expression upon sponge degradation if these bacteria leave the sponge. 3) Expression of proteins with eukaryotic-like domains, nutrient importers, presence of transposable elements and viral defense, and metabolism that could provide certain nutrients (e.g., vitamins, amino acids, organosulfur compounds) to the host, might decrease in expression as the sponge begins to disappear.

Results that may apply to this third category include the significant functions observed to decrease in expression in the genus Flavobacterium. These include: T9SS C-terminal target domain, fibronectin type III domain-containing protein, and ankyrin repeat domain. T9SS C- terminal target domain, is a type IX secretion system, only found in some species of the

Bacteroidetes phylum, and has one of two roles: a means of gliding movement for commensal or environmental bacteria or involved in virulence secretion by pathogens (Lasica et al., 2017).

Interestingly, the non-pathogenic bacteria with T9SS are known to degrade cellulose or chitin

(Lasica et al., 2017), while other secretion systems enriched in the marine sponge symbiont

Entotheonella included T2SS associated with degradation of collagen (Liu et al., 2016). The ability to degrade collagen was hypothesized to be critical for Entotheonella to create space in the sponge mesohyl (Liu et al., 2016). This is the first record of T9SS in putative sponge symbionts and may be involved in nutrient acquisition and/or interaction with the host.

Fibronectin type III domain-containing protein and, mainly, ankyrin repeat domains are enriched in sponge-symbiotic community in marine sponges compared to free-living communities (Fan et al., 2012; Fiore et al., 2015; Liu et al., 2016; Thomas et al., 2010). These eukaryotic-like domains presumably facilitate interaction with host cells or proteins, although the mechanisms for this are not well characterized. All of these traits would be expected in symbiotic bacteria and were observed to decrease in abundance in the Flavaobacterium genus in Post samples. Therefore, some of the Flavobacterium observed here may be facultative

39 symbionts that leave the host as the sponge tissue degrades and/or cued by other environmental factors.

The sponges collected for metatranscriptome analysis were collected over three months in summer and early fall where the temperature, irradiance levels, turbidity, and water levels change in addition to physiological changes in the sponge. Environmental parameters (pH, temperature, and conductivity) did not appear to vary visibly over the three months at the New

River collection site, with the exception of an increase in pH in August. Phytoplankton also increased in abundance in August at the New river site. For comparison, another location that consistently contains sponges, the Jacob Fork river, was also tested for nutrient and environmental parameters. At this particular site, we typically observe larger sponges, sponges that are green in color (harboring algal symbionts), and sponges are generally present in the water for longer into the fall and winter than at the New river site. However, the measurements of environmental parameters at Jacob Fork were generally opposite of what we expected with lower abundance of bacteria (although consistently higher abundances of phytoplankton), lower pH, lower conductivity, and similar water temperatures. Overall, irradiance levels and pH and/or changes in water level could be environmental factors affecting the microbial communities in the water and in the sponges.

The results described here support the presence of some symbiotic bacteria in the sponge and potential for obligate or at least long-term symbiosis that could be important throughout the sponge life cycle. Future work is needed to track sponge microbiome communities over time and between species and compare the phylogeny of these taxa with other freshwater sponge taxa. Additionally, analysis of the host sponge transcriptome as a next step, using the marine sponge genome, Amphimedon queenslandica, should shed light on physiological processes occurring in the sponge during gemmule formation. Another consideration is the use of short reads in the annotation analysis as conducted in the present

40 work. The annotation in the SAMSA2 pipeline relied on reads only, and a logical next step is to create a de novo assembly of the metatranscriptome to annotating the contigs, rather than the reads. Ideally, we would be able to then compare active functional annotations between the microbes and sponge. This would also provide insight into the proportions of sponge reads versus reads from microorganisms. Lastly, work is in progress to sequence 16S rRNA genes using DNA from gemmules isolated from sponges in the present study. The work presented here has also laid the foundation for future experimental work with gemmules in the laboratory to ultimately move beyond correlation in the sponge-microbe community addressed here.

Bibliography

Aguilar-Camacho JM, Doonan L, McCormack GP. 2019. Evolution of the main skeleton- forming genes in sponges (phylum Porifera) with special focus on the marine Haplosclerida (class Demospongiae). Molecular Phylogenetics and Evolution 131:245- 253 DOI 10.1016/j.ympev.2018.11.015.

Almeida EL, Rincón AFC, Jackson ST, Dobson ADW. 2019. Comparative Genomics of Marine Sponge-Derived Streptomyces spp. Isolates SM17 and SM18 With Their Closest Terrestrial Relatives Provides Novel Insights Into Environmental Niche Adaptations and Secondary Metabolite Biosynthesis Potential. Frontiers in Microbiology 10:1713 DOI 10.3389/fmicb.2019.01713.

Anwar MZ, Lanzen A, Bang-Andreasen T, Jacobsen CS. 2019. To assemble or not to resemble-A validated Comparative Metatranscriptomics Workflow (CoMW). GigaScience 8:1-10 DOI 10.1093/gigascience/giz096.

Bart MC, de Vet SJ, de Bakker DM, Alexander BE, van Oevelen D, van Loon EE, van Loon JJWA, de Goeij JM. 2019. Spiculous skeleton formation in the freshwater sponge Ephydatia fluviatilis under hyper gravity conditions. PeerJ 6:e6055:1-18 DOI 10.7717/peerj.6055.

Blunt JW, Copp BR, Munro MGH, Northcote PT, Prinsep MR. 2004. Marine natural products. Natural Produce Reports 21(1):1-49 DOI 10.1039/b305250h.

Bolger AM, Lohse M, Usadel B. 2014. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics 30(15):2114-2120.

Brestoff JR, Artis D. 2013. Commensal bacteria at the interface of host metabolism and the immune system. Nature Immunology 14:676-684.

Buchfink B, Xie C, Huson DH. 2015. Fast and sensitive protein alignment using DIAMOND. Nature Methods 12(1):59-60.

Callahan BJ, McMurdie PJ, Rosen MJ, Han AW, Johnson AJ, Holmes SP. 2016. DADA2: High resolution sample inference from Illumina amplicon data. Nature Methods 13(7):581- 583 DOI 10.1038/nmeth.3869.

Caporaso JG, Ackermann G, Apprill A, Bauer M, Berg-Lyons D, Betley J, Fierer N, Fraser L, Fuhrman JA, Gilbert JA, Gormley N, Humphrey G, Huntley J, Jansson JK, Knight R, Lauber CL, Lozupone CA, McNally S, Needham DM, Owens SM, Parada AE, Parsons R, Smith G, Thompson LR, Thompson L, Turnbaugh PJ, Walters WA, Weber L. 2018. EMP 16S Illumina Amplicon Protocol. protocols.io DOI 10.17504/protocols.io.nuudeww.

Costa R, Keller-Costa T, Gomes NCM, da Rocha UN, van Overbeek L. 2013. Evidence for selective bacterial community structuring in the freshwater sponge Ephydatia fluviatilis. Microbial Ecology 65(1):232-244 DOI 10.1007/s00248-012-0102-2.

42 de Bary A. 1879. Die Erscheinung der Symbiose. Strasburg: Triibner.

Douglas AE. 2011. Lessons from Studying Insect Symbioses. Cell Host Microbe 10(4):359-367 DOI 10.1016/j.chom.2011.09.001.

Dreyer J, Malan AP, Dicks LMT. 2018. Bacteria of the Genus Xenorhabdus, a Novel Source of Bioactive Compounds. Frontiers in Microbiology 9:3177 DOI 10.3389/fmicb.2018.03177.

Eerkes-Medrano D, Feehan CJ, Leys SP. 2014. Sponge cell aggregation: checkpoints in development indicate a high level of organismal complexity. Invertebrate Biology 134(1):1-18 DOI 10.111/ivb.12072.

Fan L, Reynolds D, Liu M, Stark M, Kjellelberg S, Webster NS, Thomas T. 2012. Functional equivalence and evolutionary convergence in complex communities of microbial sponge symbionts. PNAS 109(27):E1878-E1887 DOI 10.1073/pnas.1203287109.

Fiore CL, Labrie M, Jarett JK, Lesser MP. 2015. Transcriptional activity of the giant barrel sponge, Xestospongia muta Holobiont: molecular evidence for metabolic interchange. Frontiers in Microbiology 6(364):2-4 DOI 10.3389/fmicb.2015.00364.

Freeman CJ, Easson CG. 2016. Sponge distribution and the presence of photosymbionts in Moorea, French Polynesia. PeerJ 4:e1816 DOI 10.7717/peerj.1816

Freeman CJ, Thacker RW. 2011. Complex interactions between marine sponges and their symbiotic microbial communities. Limnology and Oceanography. 56(5):1577-1586 DOI 10.4319/lo.2011.56.5.1577.

Frost TM. 1978. The impact of freshwater sponge Spongilla lacustris on sphagnum bog-pond. Verhandlungen Internationale Vereinigung für Theoretische und Angewandte Limnologie 20(4):2368-2371.

Frost TM, Williamson CE. 1980. In situ determination of the effect of symbiotic algae on the growth of the freshwater sponge Spongilla lacustris. Ecology 61:1361-1370.

Funayama N, Nakatsukasa M, Hayashi T, Kiyokazu A. 2005. Isolation of the choanocyte in the fresh water sponge, Ephydatia fluviatilis and its lineage marker, Ef annexin. Development, Growth and Differentiation 47(4):243-253 DOI 10.1111/j.1440- 169X.2005.00800.x.

Gaikwad S, Shouche YS, Gade WN. 2016. Microbial community structure of two freshwater sponges using Illumina MiSeq sequencing revealed high microbial diversity. AMB Express 6:40 DOI 10.1186/S13568-016-0211-2.

Gernert C, Glockner FO, Krohne G, Hentschel U. 2005. Microbial diversity of the freshwater sponge Spongilla lacustris. Microbial Ecology 50(2):206-212 DOI 10.1007/s00248-004- 0172-x.

Gilbert JJ, Simpson TL. 1976. Sex reversal in a freshwater sponge. Journal of Experimental Zoology 195(1):145-151.

Gladkikh AS, Kalyuzhnaya OV, Belykh OI, Ahn TS, Parfenova VV. 2014. Analysis of bacterial communities of two Lake Baikal endemic sponge species. Mikrobiologiia 83(6):787–97.

Grabherr M, Haas BJ, Yassour M, Levin JZ, Thompson DA, Amt I, Adiconis X, Fan L, Raychowdhury R, Zeng Q, Chen Z, Mauceli E, Hacohen N, Gnirke A, Rhind N, di Palma F, Birren BW, Nusbaum C, Lindblad-Toh K, Friedman N, Regev A. 2011. Full- length transcriptome assembly from RNA-Seq data without a reference genome. Nature Biotechnology 29(7):644-652 DOI 10.1038/nbt.1883.

Gunduz EA, Douglas AE. 2009. Symbiotic bacteria enable insect to use a nutritionally inadequate diet. Proceedings of the Royal Society B-Biological Sciences 276:987–991.

Guzii AG, Makarieva TN, Denisenko VA, Dmitrenok PS, Kuzmich AS, Dysholovoy SA, Krasokhin VB, Stonik VA. 2010. Monanchocidin: a new apoptosis-inducing polycyclic guanidine alkaloid form the marine sponge Monanchora pulchra. Organic Letters 12(19):4292-4295 DOI 10.1021/ol101716x.

Harrison FW. 1974. Sponges (Porifera: Spongillidae). In: Hart CW & Fuller SLH, ed. Pollution Ecology of Freshwater Invertebrates. New York: Academic Press, 29-66.

Harrison FW, Johnston L, Stansell KB, McAndrew W. 1977. The taxonomic and ecological status of the environmentally restricted spongillid species of North America. I. Spongilla sponginosa Penney 1957. Hydrobiologia 53:199-202.

Harsha RE, Francis JC, Poirrier MA. 1983. Water temperature: a factor in the seasonality of two freshwater sponge species, Ephydatia fluviatilis and Spongilla alba. Hydrobiologia 102(3):145-150 DOI 10.1007/bf00006340.

Heckmann K, Schmidt HJ. 1987. Polynucleobacter necessarius gen. nov., sp. nov., an Obligately Endosymbiotic Bacterium Living in the Cytoplasm of Euplotes aediculatus. International Journal of Systematic and Evolutionary Microbiology 37(4):456–457.

Hentschel U, Hopke J, Horn M, Friedrich AB, Wagner M, Hacker J, Moore BS. 2002. Molecular Evidence for a Uniform Microbial Community in Sponges from Different Oceans. Applied and Environmental Microbiology 68(9):4431-4440 DOI 10.1128/AEM.68.9.4431-4440.2002.

Hentschel U, Fieseler L, Wehrl M, Gernet C, Steinert M, Hacker J, Horn M. 2003. Microbial diversity of marine sponges. In: Muller, WEG (Ed.) Molecular Marine Biology of Sponges. Springer, Berlin, Heidelberg, New York:60-88.

Hentschel U, Piel J, Degnan SM, Taylor MW. 2012. Genomic insights into the marine sponge microbiome. Nature Reviews 10: 641-654.

Hill A, Schenkelaars Q, Hall C, Fierro L, Renard E, Borchiellini C. Unpublished-2014. Pick, Wash and Hatch Freshwater Sponge Gemmules Handbook. University of Richmond, Department of Biology.

Hoetzinger M, Schmidt J, Pitt A, Koll U, Lang E, Hahn MW. 2018. Polynucleobacter paneuropaeus sp. nov., characterized by six strains isolated from freshwater lakes located along a 3000 km north–south cross-section across Europe. International Journal of Systematic and Evolutionary Microbiology 69(1):203-213 DOI 10.1099/ijsem.0.003130.

Hoffmann F, Radax R, Woebken D, Holtappels M, Lavik G, Rapp HT, Schläppy ML, Schleper C, Kuypers MMM. 2009. Complex nitrogen cycling in the sponge Geodia barretti. Environmental Microbiology 11(9):2228-2243 DOI 10.111/j.1462-2920.2009.01944.x.

Höhr D. 1977. Differenzierungsvorgänge in der keimenden Gemmula vonEphydatia fluviatilis. Wilhelm Roux’s Archives of Developmental Biology 182(4):329-346 DOI 10.1007/bf00848384.

Hooper JN, Van Soest RWM. 2002. Systema Porifera. A Guide to the Classification of Sponges. In: Hooper J & van Soest RWM, eds. System Porifera: A Guide to the Classification of Sponges. New York: Kluwer Academic/Plenum Publishers, 1-7.

Horn H, Slaby BM, Jahn MT, Bayer K, Moitinho-Silva L, Förster F, Abdelmohsen UR, Hentschel U. 2016. An Enrichment of CRISPR and Other Defense-Related Features in Marine Sponge-Associated Microbial Metagenomes. Frontiers in Microbiology 7:1751 DOI 10.3389/fmicb.2016.01751.

Ilan M, Dembo G, Gasith A. 1996. Gemmules of sponges from warm lake. Freshwater Biology 35(1):165-172 DOI 10.1046/j.1365-2427.1996.00486.x.

Imsiecke G, Steffen R, Custodio M, Borojevic R, Muller WG. 1995. Formation of Spicules by Sclerocytes from the Freshwater Sponge Ephydatia Muelleri in Short-Term Cultures In Vitro. In Vitro Cellular & Developmental Biology-Animal 131:528-535.

Kamke J, Sczyrba A, Ivanova N, Schwientek P, Rinke C, Mavromatis K, Woyke T, Hentschel U. 2013. Single-cell genomics reveals complex carbohydrate degradation patterns in poribacterial symbionts of marine sponges. ISME Journal 7:2287-2300.

Keller-Costa T, Jousset A, van Overbeek L, van Elsas JD, Costa R. 2014. The Freshwater Sponge Ephdatia fluviatilis harbours diverse Pseudomonas species (Gammaproteobacteria, Pseudomonadales) with broad-spectrum antimicrobial activity. PLoS ONE 9(2):e88429 DOI 10.1371/journal.pone.0088429.

Kiran GS, Sivasankari S, Ramasamy P, Thinesh T, Hassan S, Lipton AN, Ninawe AS, Selvin J. 2018. Marine sponge microbial association: Towards disclosing unique symbiotic interactions. Marine Environmental Research 140:169-179 DOI 10.1016/j.marenvres.2018.04.017.

Kopylova E, Noe L, Touzet H. 2012. SortMeRNA: fast and accurate filtering of ribosomal RNAs in metatranscriptomic data. Bioinformatics 28(24):3211-3217.

Lasica AM, Ksiazek M, Madej M, Potempa J. 2017. The Type IX Secretion System (T9SS): Highlights and recent insights into its structure and function. Frontiers in Cellular and Infection Microbiology 7:215 DOI 10.3389/fcimb.2017.00215.

Leimena MM, Ramiro-Garcia J, Davids M, van den Bogert B, Smidt H, Smid EJ, Boekhorst J, Zoetendal EG, Schaap PJ, Kleerebezem M. 2013. A comprehensive metatranscriptome analysis pipeline and its validation using human small intestine microbiota datasets. BMC Genomics 14:530 DOI 10.1186/1471-2164-14-530.

Liu MY, Kjelleberg S, Thomas T. 2011. Functional genomic analysis of an uncultured δ- proteobacterium in the sponge Cymbastela concentrica. ISME Journal 5:427-435 DOI 10.1038/ismej.2010.139.

Liu F, Li J, Feng G, Li Z. 2016. New Genomic Insights into “Entotheonella” Symbionts in Theonella swinhoei: Mixotrophy, Anaerobic Adaptation, Resilience, and Interaction. Frontiers Microbiology 7:Article 1333 DOI 10.3389/fmicb.2016.01333.

Loh TL, Pawlik JR. 2014. Chemical defenses and resource trade-offs structure sponge communities on Caribbean coral reefs. PNAS 111(11): 4151-4156 DOI 10.1073/pnas.1321626111.

Lloyd-Price J, Abu-Ali G, Huttenhower C. 2016. The healthy human microbiome. Genome Medicine 8:51 DOI 10.1186/s13073-016-0307-y.

Loomis SH, Bettridge A, Branchini BR. 2008. The effects of elevated osmotic concentration on control of germination in the gemmules of freshwater sponges Eunapius fragilis and Anheteromeyania ryderi. Physiological and Biochemical Zoology 82(4): 388-395 DOI 10.1086/589901.

Love GD, Grosjean E, Stalvies C, Fike DA, Grotzinger JP, Bradley AS, Kelly AE, Bhatia M, Meredith W, Snape CE, Bowring SA, Condon DJ, Summons RE. 2009. Fossil steroids record the appearance of Demospongiae during the Cryogenian period. Nature 456(7230):718-721 DOI 10.1038/nature07673.

Love MI, Huber W, Anders S. 2014. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biology 15:550 DOI 10.1186/s1305-014-0550-8.

Luo W, Brouwer C. 2013. Pathview: an R/Bioconductor package for pathway-based data integration and visualization. Bioinformatics 29(14):1830-1831 DOI 10.1093/bioinformatics/btt285.

Manconi R, Prozato R. 2002. Suborder Spongillina subord. nov.: Freshwater sponges. In: Hooper J & van Soest RWM, eds. System Porifera: A Guide to the Classification of Sponges. New York: Kluwer Academic/Plenum Publishers, 921-1019.

Manconi R, Pronzato R. 2007. Gemmules as a key structure for the adaptive radiation of freshwater sponges: a morphofunctional and biogeographical study. In: Custodia MR, Hajdu E, Lobo-Hadju G, Muricy G, eds. Porifera Research: Biodiversity, Innovation and Sustainability. Proceedings of the 7th International Symposium 2006:61-77.

Manconi R, Pronzato R. 2008. Global distribution of sponges (Porifera: Spongillidae) in freshwater. Hydrobiologia. 595(1): 27-33 DOI 10.1007/s10750-007-9000-x.

Mayer AM, Glaser KB, Cuevas C, Jacobs RS, Kem W, Little RD, McIntosh JM, Newman DJ, Potts BC, Shuster DE. 2010. The odyssey of marine pharmaceuticals: a current pipeline perspective. Trends Pharmacological Science 31(6):255-265 DOI 10.1016/j.tips.2010.02.005.

McFall-Ngai MJ, Ruby EG. 1998. Sepiolids and vibrios: when they first meet. BioScience 48:257–265.

McMurdie PJ, Holmes S. 2013. phyloseq: An R Package for Reproducible Interactive Analysis and Graphics of Microbiome Census Data. PLoS ONE 8(4): e6121. DOI 10.1371/journal.pone.0061217.

McMurdie PJ, Holmes S. 2014. Waste Not, Want Not: Why Rarefying Microbiome Data is Inadmissible. PLoS Computational Biology 10(4):e1003531 DOI 10.1371/journal.pcbi.1003531.

Mohamed NM, Saito K, Tal Y, Hill RT. 2010. Diversity of aerobic and anaerobic ammonia- oxidizing bacteria in marine sponges. ISME Journal 4(1):38-48 DOI 10.1038/ismej.2009.84.

Moitinho-Silva L, Díez-Vives C, Batani G, Esteves AI, Jahn MT, Thomas T. 2017. Integrated metabolism in sponge-microbe symbiosis revealed by genome-centered Metatranscriptomics. ISME Journal 11:1651-1666.

Morrow C, Cárdenas P. 2015. Proposal for a revised classification of the Demospongiae (Porifera). Frontiers in Zoology 12:7 DOI 10.1186/s12983-015-0099-8.

Nakayama S, Arima K, Mohri K, Funayama N. 2010. Spiculous skeleton formation as a new model to clarify pattern formation in demosponges: the roughly spaced spicule holding up (SHU) points and the identification of spicule carrying cells. In: Proceedings of 8th World Sponge Symposium, Girona, 86.

Nyholm SV, McFall-Ngai MJ. 2004. The Winnowing: Establishing The Squid-Vibrio Symbiosis. Nature Reviews Microbiology 2(8):632-42 DOI 10.1038/nrmicro 957.

Olson JB, Gochfeld DJ, Slattery M. 2006. Aplysina Red Band Syndrome: A New Threat to Caribbean Sponges. Diseases of Aquatic Organisms 71(2):163-168 DOI: 10.3354/dao071163.

Overbeek R, Olson R, Pusch GD, Olsen GJ, Davis JJ, Disz T, Edwards RA, Gerdes S, Parrello B, Shukla M, Vonstein V, Wattam AR, Xia F, Stevens R. 2014. The SEED and the Rapid Annotation of microbial genomes using Subsystems Technology (RAST). Nucleic Acids Research 42(Database issue):D206-D214 DOI 10.1093/nar/gkt1226.

Penney JT, Racek AA. 1968. Comprehensive revision of a worldwide collection of freshwater sponges (Porifera: Spongillidae). Bulletin of the United States National Museum 272:1- 184 DOI 10.549/si.03629236.272. Petersen JM, Osvatic J. 2018. Microbiomes in natura: importance of invertebrates in understanding the natural variety of animal-microbe interactions. mSystems 3(2):e00179-17

Pronzato R, Manconi R. 2002. Atlas of European Freshwater Sponges. Annali del Museo Civico di Storia naturale di Ferrare 4:3-64.

Quah Y, Ismail NIM, Ooi JLS, Affendi YA, Manan FA, Wong FC, Chai TT. 2017. Identification of novel cytotoxic peptide KENPVLSLVNGMF from marine sponge Xestospongia testudinaria, with characterization of stability in human serum. International Journal of Peptide Research and Therapeutics 24(1):189-199 DOI 10.1007/s10989-017-9604-6.

Reveillaud J, Maignien L, Eren MA, Huber JA, Apprill A, Sogin ML, Vanreusel A. 2014. Host- specificity among abundant and rare taxa in the sponge microbiome. ISME Journal 8(6):1198-1209 DOI 10.1038/ismej.2013.227.

Ricciardi A, Resiwig HM. 1992. Freshwater sponges (Porifera, Spongillidae) of eastern Canada: taxonomy, distribution, and ecology. Canadian Journal of Zoology 71:665-682.

Rozenfeld F. 1970. Inhibition du developpement des gemmules de spongillides: specificite et moment d’action de la gemmulostasine. Archives de Biologie Liege 81:193-214.

Ruthman A. 1965. The fine structure of RNA-storing archaeocytes from gemmules of freshwater sponges. Quarterly Journal of Microscopical Science 106:99-114.

Sacristan-Soriano O, Banaigs B, Casamayor EO, Becerro MA. 2011. Exploring the links between natural products and bacterial assemblages in the sponge Aplysina aerophoba. Applied and Environmental Microbiology 77(3):862-870 DOI 10.1128/AEM.00100-10.

Sayers EW, Barrett T, Benson DA, Bolton E, Bryant SH, Canese K, Chetvernin V, Church DM, Dicuccio M, Federhen S, Feolo M, Geer LY, Helmberg W, Kapustin Y, Landsman D, Lipman DJ, Lu Z, Madden TL, Madej T, Maglott DR, Marchler-Bauer A, Miller V, Mizrachi I, Ostell J, Panchenko A, Pruitt KD, Schuler GD, Sequeira E, Sherry ST, Shumway M, Sirotkin K, Slotta D, Souvorov A, Starchenko G, Tatusova TA, Wagner L, Wang Y, John Wilbur W, Yaschenko E, Ye J. 2010. Database resources of the National Center for Biotechnology Information. Nucleic Acids Research 38:D5-D16 DOI 10.1093/nar/gkp967.

Schmitt S, Wehrl M, Bayer K, Siegl A, Hentschel U. 2007. Review: Marine sponges as models for commensal microbe-host interactions. Symbiosis 44:43-50.

Schmitt S, Tsai P, Bell J, Fromont J, Ilan M, Lindquist N, Perez T, Rodrigo A, Schupp PJ, Vacelet J, Webster N, Hentschel U, Taylor MW. 2012. Assessing the complex sponge microbiota: core, variable and species-specific bacterial communities in marine sponges. ISME Journal 6(3):564-576.

Seipke RF, Kaltenpoth M, Hutchings MI. 2012. Review: Streptomyces as symbionts: an emerging and widespread theme? FEMS Microbiology Reviews 36(4):862-876.

Seutin G, White BN, Boag PT. 1991. Preservation of avian blood and tissue samples for DNA analyses. Canadian Journal of Zoology 69(1):82-90 DOI 10.1139/z91-013.

Sharp K, Earm B, Faulkner D, Haygood M. 2007. Vertical transmission of diverse microbes in the tropical sponge Corticium sp. Applied Environmental Microbiology 73:622-629.

Simpson TL. 1984. The Cell Biology of Sponges. New York; Springer-Veriage.

Simpson TL, Fell PE. 1974. Dormancy among the Porifera: Gemmule Formation and Germination in Fresh-Water and Marine Sponges. Transactions of the American Microscopical Society 93(4):544-577.

Simpson TL, Vaccaro CA. 1974. An ultrastructural study of silica deposition in the freshwater sponge Spongilla lacustris. Journal of Ultrastructure. Research 47:296-309.

Slaby BM, Hackl T, Horn H, Bayer K, Hentschel U. 2017. Metagenomic binning of a marine sponge microbiome reveals unity in defense but metabolic specialization. ISME Journal 11:2465-2478.

Steger D, Ettinger-Epstein P, Whalan S, Hentschel U, de Nys R, Wagner M, Taylor MW. 2008. Diversity and mode of transmission of ammonia-oxidizing archaea in marine sponges. Environmental Microbiology 10:1087-1094.

Strekal TS, McDiffett WF. 1974. Factors Affecting Germination, Growth, and Distribution of the Freshwater Sponge, Spongilla Fragilis Leidy (Porifera) Biology Bulletin 146:267-278.

Tatusova T, Ciufo S, Fedorov B, O’Neill K, Tolstoy I. 2014. RefSeq microbial genomes database: new representation and annotation strategy. Nucleic Acids Research 42(Database issue):D553-559.

Taylor MW, Radax R, Steger D, Wagner M. 2007. Sponge-associated microorganisms: evolution, ecology, and biotechnological potential. Microbiology and Molecular Biology Reviews 71:259-347.

Thomas T, Moitinho-Silva L, Lurgi M, Björk JR, Easson C, Astudillo-García C, Olson JB, Erwin PM, López-Legentil S, Luter H, Chaves-Fonnegra A, Costa R, Schupp PJ, Steindler L, Erpenbeck D, Gilbert J, Knight R, Ackermann G, Victor Lopez J, Taylor MW, Thacker RW, Montoya JM, Hentschel U, Webster NS. 2016. Diversity, structure and convergent evolution of the global sponge microbiome. Nature Communications 7:11870.

Ul-Hasan S, Bowers RM, Figueroa-Montiel A, Licea-Navarro AF, Beman JM, Woyke T, Nobile CJ. 2019. Community ecology across bacteria, archaea and microbial eukaryotes in the sediment and seawater of coastal Puerto Nuevo, Baja California. PLoS ONE 14(2): e0212355 DOI 10.1371/journal.pone.0212355

Vacelet J, Donadey C. 1977. Electron microscope study of the association between some sponges and bacteria. Journal of Experimental Marine Biology and Ecology 30:301-314.

Van Weel PB. 1949. On the physiology of the tropical fresh-water sponge Spongilla proliferans (Annand). I. Ingestion, digestion and excretion. Physiologia Comparata et Ocologia; International Journal of Comparative Physiology and Ecology 1: 110-128.

Webster JM, Chen G, Hu K, Li J. 2002. Bacterial metabolites. Entomopathogenic Nematology ed. Gaugler R. (London: CAB International): 99-114 DOI 10.1079/9780851995670.0099.

Webster NS, Taylor MW, Behnam F, Lücker S, Rattei T, Whalan S, Horn M, Wagner M. 2010. Deep sequencing reveals exceptional diversity and modes of transmission for bacterial sponge symbionts. Environmental Microbiology 12(8):2070-2082 DOI 10.1111/j.1462-2920.2009.02065.x.

Westreich ST, Treiber ML, Mills DA, Korf I, Lemay DG. 2018. SAMSA2: a standalone metatranscriptome analysis pipeline. BMC Bioinformatics 19:175 DOI 10.1186/s12859- 018-2189-z.

Wilkinson CR. 1978. Microbial associations in sponges: III. Ultrastructure of the in-situ associations in coral reef sponges. Marine Biology 49: 77-185.

Wilkinson CR. 1980. Nutrient translocation from green algal symbionts to the freshwater sponge Ephydatia fluviatilis. Hydrobiologia 75(3):241-250 DOI 10.1007/bf00006488.

Wilkinson C, Fay P. 1979. Nitrogen fixation in coral reef sponges with symbiotic cyanobacteria. Nature 279:527-529.

Williamson CE. 1979. An ultrastructural investigation of algal symbiosis in white and green Spongilla lacustris (L.) (Porifera:Spongillidae). Transaction American Microscopical Society 98:59-77.

Wilson MC, Mori T, Ruckert C, Uria AR, Helf MJ, Takada K, Gernet C, Steffens UAE, Heycke N, Schmitt S, Rinke C, Helfrich EJN, Brachmann AO, Gurgui C, Wakimoto T, Kracht M, Crusemann M, Hentschel U, Abe Ikuro, Matsunaga S, Kalinowski J, Takeyama H, Piel J. 2014. An environmental bacterial taxon with a large and distinct metabolic repertoire. Nature 506:58-62 DOI 10.1038/nature12959.

Worden A. Unpublished 2009. DNA Extraction-CTAB Method. MBARI.

Zhang J, Kobert K, Flouri T, Stamatakis A. 2014. PEAR: a fast and accurate Illumina paired- end reAd mergeR. Bioinformatics (Oxford, England) 30(5):614-620.

Zhang W, Xue S, Zhao Q, Zhang X, Li J, Jin M, Yu X, Yuan Q. 2003. Biopotentials of marine sponges from China oceans: past and future. Biomolecular Engineering 20(4-6):413-419 DOI 10.1016/S1389-0344(03)00066-2.

Tables and Figures

Figure 1: Life cycle of Ephydatia fluviatilis in asexual and sexual reproduction. (A) Gemmules formed within the parent sponge tissue. (B) Parent tissue is destroyed, gemmule is exposed and followed by subsequent hatching; germination occurs. (C) cells break out of the gemmule coating. (D) These cells form a small, fully functioning sponge in about 1 week. Sexual reproduction sperm and eggs are formed within male and female sponges. (E) Eggs are fertilized by the sperm in the female sponges and begin to cleavage. (F) Larvae leave parent sponge. (G) Larvae settle and start to metamorphize. (H) A new sponge has formed (Funayama et al., 2005).

Figure 2: Developmental stages of germination in the freshwater sponge Ephydatia fluviatilis. (A) Stage

0: resting gemmule. (B) Stage 1: (2-day-old sponge) cells actively migrate from the gemmule and differentiate into epithelial cells to form an envelope. (C) Stage 2: (3-4-day old sponge) cells move into the epithelial envelope to proliferate and start differentiating into several different types of cells. (D)

Stage 3: (5-day-old sponge) the aquiferous system and choanocyte chambers are formed. (E) Stage 4: (6-

7-day-old sponge) sponge is fully functionable with the formation of the oscula in this miniature sponge.

(F) Stage 5: (8-12-day-old sponge) sponge maintains its current body form and gradually increases in size from this point on (Funayama et al., 2005).

Figure 3: Vertical sectioned view of a hatching sponge from a gemmule. A, archeocyte; Bp, basopinacocyte; Ca, water canal; Ch, choanocyte; ChC choanocyte chamber; Enp, endopinacocyte; Exp, exopinacocyte; GC, gemmule coat; L, lophocyte, Co, collagenocyte; Oc, oscle; Sc, sclerocyte; Sp, spicule

(Funayama et al., 2005).

Figure 4: Modified SAMSA2 pipeline flow chart to include all the steps taken during the sequence assembly and data analysis process (Westreich et al., 2018).

Figure 5: Bar graph of relative abundance of taxonomic hits at class level on Y-axis with sample and collection month information on X-axis. Sponge samples are on the left side of the chart, while water samples are on the right side. Phylum associated with each of the classes is listed on the key in parentheses.

Figure 6: The network representation of the similarity in microbiome profile between the sponges and river water samples collected for preliminary data. Network is based on Bray-Curtis distance, length and thickness of lines indicate similarity between samples. The thicker and shorter the lines are the more similar the samples. Two separate networks were detected here with July samples (green dots) for water and sponge different from the August and September (red and blue) samples, indicating they are significantly different.

Figure 7: Heat map displaying the abundance of the top 300 ASVs at the phylum level for all sponge and water samples collected for preliminary data. The red oval represents ASVs that may be enriched or specific to sponges.

Class *Chitinophagales Cytophagales

Flavobacteriales Sphingobacteriales

Figure 8: Heat map displaying the abundance of the ASVs among the phylum Bacteroidetes for all sponge and water samples. The red oval represents ASVs that may be enriched or specific to sponges. The classes that the circled ASVs comprise are listed with an asterisk marking the most abundant.

Figure 9: PCA plot of Pre (blue) and Post (red) Samples by organismal data, with the Y-axis, PC2, with a

22% variance and the X-axis, PC1, with a 54% variance.

Figure 10: PCA plot of Pre (blue) and Post (red) Samples by functional data, with the Y-axis, PC2, with a

18% variance and the X-axis, PC1, with a 57% variance.

total sample total

Relative activity of of Relativeactivity

16 17 50 55 58 24 25 57

readssampleper

Total

16 17 50 55 58 24 25 57

Figure 11: Relative activity graph of the top 10 functional data for each sample. The top bar graph is of

relative activity of total sample while the bottom bar graph is total reads per sample. The samples are

divided into Pre (left) and Post (right) with a color key located at the bottom.

Figure 12: Pie chart representing the top 10 functional subsystems represented in the metatranscriptome samples by SEED.

Bacteria

Phytoplankton

3.00E+06 3.00E+04 Phytoplanktoncell (cells/mL) density *

2.50E+06 * 2.50E+04 * * * 2. 00E+06 2.00E+04 * *

1.50E+06 1.50E+04

1.00E+06 1.00E+04

5. 00E+05 5.00E+03 Bacterial cell density (cells/mL) density cell Bacterial

0. 00E+00 0.00E+00 9 9 9 9 9 9 1 19 1 1 1 1 1 0 0 0 0 0 0 0 2 2 2 2 2 t2 2 ly g t y g p _ u u ep l u e y J A S Ju A S a ______M JF F F R R R _ J J N N N d on P Bacteria Series2 Series3 Pico/Nanoplankton

Figure 13: A bar chart showing the cell density (cells/mL) of Bacteria (purple) and Phytoplankton

(green) via flow cytometry in water samples collected at Jacob Fork river (JF) and New river (NR) with

corresponding months and years of collection.

Figure 14: A series of bar charts showing different environmental factors collected on the Y-axis for each site, New river (grey) and Jacob Fork river (black), with the corresponding months (July-October) on the

X-axis. From top to bottom: pH, temperature, and conductivity.

Figure 15: A series of bar charts showing different nutrient concentrations (mg L-1) in collected water samples on the Y-axis, from top to bottom: Chloride, Nitrate, and Sulfate with the corresponding collection months and locations on the X-axis with NR-New river and JF-Jacob Fork river. The collection years are represented by grey bars (2017) and black bars (2018).

Table 1: Major cell types found in gemmules and mature sponge bodies

Cell Types Function

Archeocytes Cellular differentiation; Role in nutrient transport; Ingestion and digestion of food particles; Role in sexual reproduction

Thesocytes Resting state of archeocytes in the gemmule

Choanocytes Collar cells; Generate water flow

Sclerocytes Spicule making cells

Exopinacocytes Outer epithelial cells

Endopinacocytes Inner epithelial cells

Basopinacocytes Basal epithelial cells; Positions and anchors sclerocytes

Spongocytes Active in collagen biosynthesis

Collagenocytes Active in collagen biosynthesis

Lophocytes Active in collagen biosynthesis

Pinacocytes Pavement cells; Encloses the mesohyl

Transport cells Transport sclerocytes to their location

Table 2: Qualitative data for each of the samples used in the study, such as the sample ID number, year and month in which they were collected, pre or post gemmule formation, and the location of collection site.

Sample_ID Year Pre_Post Month Collection_site

16 2018 Pre July New river 17 2018 Pre July New river

24 2018 Post August New river 25 2018 Post August New river

50 2019 Pre July New river 55 2019 Pre August New river 57 2019 Post September New river 58 2019 Pre September New river

Table 3: The loss of 16S reads during the trimming process. The original raw reads for each sample, read counts after filtering, read counts after denoised for forward and reverse reads, and nonchimeric reads.

Sample_ Input_Reads Filtered_ Denoised_F Denoised_R Merged_Reads Nonchim_ ID Reads Reads CF11 46585 10769 10033 10098 7902 7813 CF12 88385 32441 29358 29373 20170 19716 CF13 78567 36294 33289 32890 21273 20795 CF14 73395 33401 32323 32284 22128 21414 CF15 50258 24180 23074 23285 17389 16958 CF16 65129 29235 28044 27888 20855 20448 CF17 76310 31211 30081 29671 21821 21292 CF18 95249 32067 30875 30576 20519 20117 CF19 71818 34712 33570 33431 21507 20739 CF20 77343 31184 30237 29937 19597 18719 CF21 36194 18858 18159 18126 11248 11069 CF22 38926 19580 18563 18575 13852 13751 CF23 47019 22044 20863 20721 14114 13744 CF24 52029 27632 26537 26424 16460 15928 CF26 69827 31118 29694 29419 19776 19367 CF27 69628 31285 29731 29559 19847 19401 CF28 59333 33011 31211 31041 18910 18295 CF29 76080 43502 42035 41900 24824 24089 CF30 76211 35093 33800 33530 21338 20644 CF31 46890 23981 22771 22227 11606 11310 CF32 76719 38033 36654 36571 22344 21970 CF40 45830 23656 22731 22651 12489 12103 CF41 79619 50765 49599 49389 19250 17635 CF42 38340 20866 20059 20019 11176 10700 CF43 76839 30794 28804 28888 20439 19581 CF44 76391 34319 32619 32776 23599 22353 CF45 62189 22116 22116 21978 15526 14520

Table 4: The loss of RNA reads during the quality trimming process. The original raw reads for each sample, read counts after PEAR merged paired reads, and the final read count after Trimmomatic removed low-quality reads.

Sample_ID Raw_Reads Post_Merge Final_Read_Count 16 47,708,555 39,059,383 35,517,425 17 52,779,588 43,326,124 39,353,480 24 37,446,129 30,297,596 27,372,921 25 50,833,021 41,473,488 37,731,589 50 54,321,279 44,313,614 40,118,836 55 46,919,546 35,739,660 32,703,117 57 52,442,778 44,851,409 40,647,766 58 41,752,015 38,174,036 33,097,311

Table 5: Index table of Shannon and Simpson diversity between Pre and Post gemmule samples for organisms and functional results.

Organism Results Pre Post

Shannon diversity index 6.12 6.18

Simpson diversity index 0.98 0.98

Functional Results

Pre Post

Shannon diversity index 4.65 4.69 Simpson diversity index 0.87 0.88

Table 6: Organisms that were significantly different in abundance between the pre and post gemmule

sponge samples. Differential abundance was calculated using DESeq2 and the table includes mean (X),

log2 fold change, indicating if the organisms increased or decreased in abundance, the error associated

with the fold change (lfcSE), as well as the statistic (stat), p-value, and adjusted p-value.

Organism Name baseMean PreMean PostMean log2Fold lfcSE stat pvalue padj Change Kitasatospora cheerisanensis 178.35 369.44 63.69 -2.53 0.34 -7.39 1.49E-13 1.10E-09 Flavobacterium psychrophilum 767.79 528.37 911.44 0.79 0.13 5.99 2.12E-09 7.82E-06 Flavobacterium cauense 353.09 224.50 430.24 0.95 0.16 5.76 8.60E-09 2.11E-05 Bacillus soli 188.34 400.60 60.98 -2.72 0.54 -5.00 5.86E-07 1.08E-03 Sphingomonas melonis 97.10 36.45 133.50 1.86 0.40 4.60 4.25E-06 6.27E-03 Polynucleobacter asymbioticus 2290.98 3348.48 1656.48 -1.01 0.23 -4.49 7.09E-06 8.72E-03 Streptomyces milbemycinicus 47.85 75.13 31.49 -1.28 0.29 -4.41 1.05E-05 1.11E-02 Flavobacterium filum 702.96 461.86 847.62 0.88 0.20 4.37 1.23E-05 1.14E-02 Polynucleobacter sp. 22264.68 31491.05 16728.86 -0.91 0.21 -4.28 1.88E-05 1.54E-02 Pseudoglutamicibacter albus 29.30 55.73 13.44 -2.04 0.48 -4.24 2.20E-05 1.62E-02 Flavobacterium aquatile 3146.72 1701.89 4013.62 1.24 0.30 4.16 3.15E-05 1.92E-02 Polynucleobacter necessarius 1174.49 1691.93 864.03 -0.97 0.24 -4.07 4.65E-05 1.92E-02 Flavobacterium sasangense 603.10 363.48 746.88 1.04 0.25 4.10 4.12E-05 1.92E-02 Flavobacterium suncheonense 363.99 271.51 419.47 0.63 0.15 4.14 3.41E-05 1.92E-02 Sphingomonas taxi 298.17 179.16 369.57 1.04 0.25 4.18 2.95E-05 1.92E-02 Promicromonosporaceae bacterium 269.82 384.84 200.81 -0.93 0.23 -4.07 4.69E-05 1.92E-02 Sulfurimonas denitrificans 209.86 271.01 173.16 -0.64 0.16 -4.10 4.06E-05 1.92E-02 Pseudoxanthomonas spadix 99.50 56.16 125.50 1.17 0.29 4.10 4.18E-05 1.92E-02 Polynucleobacter wuianus 2730.88 3992.19 1974.10 -1.02 0.25 -4.01 6.06E-05 2.23E-02 Herbidospora daliensis 72.19 102.90 53.76 -0.92 0.23 -4.02 5.79E-05 2.23E-02 Flavobacterium saliperosum 319.35 229.11 373.50 0.71 0.18 3.98 6.78E-05 2.38E-02 Polynucleobacter 3885.36 5615.76 2847.12 -0.98 0.25 -3.97 7.28E-05 2.44E-02 Phormidium sp. 955.29 1132.17 849.16 -0.42 0.11 -3.89 1.02E-04 3.27E-02 Polynucleobacter sinensis 1148.34 1581.13 888.66 -0.83 0.22 -3.84 1.23E-04 3.77E-02 Unclassified Nitrosomonadales 530.29 750.10 398.40 -0.91 0.24 -3.83 1.31E-04 3.85E-02 Xenorhabdus khoisanae 76.55 179.38 14.86 -3.58 0.95 -3.78 1.60E-04 4.54E-02

Table 7: Functions that were significantly different in abundance between the pre and post gemmule

sponge samples. Differential abundance was calculated using DESeq2 and the table includes mean (X),

log2 fold change, indicating if the functions increased or decreased in abundance, the error associated

with the fold change (lfcSE), as well as the statistic (stat), p-value, and adjusted p-value.

Functions baseMean PreMean PostMean log2Fold lfcSE stat pvalue padj Change uroporphyrin-III methyltransferase 38.07 88.18 8.00 -3.42 0.45 -7.54 4.59E-14 3.40E-10 catechol 2,3-dioxygenase 330.24 141.29 443.61 1.65 0.27 6.09 1.11E-09 4.11E-06 bifunctional o-acetylhomoserine/ o-acetylserine sulfhydrylase 70.63 118.03 42.20 -1.49 0.29 -5.05 4.43E-07 1.09E-03 porin family protein 255.87 156.20 315.66 1.02 0.22 4.71 2.47E-06 4.58E-03 DUF4293 domain-containing protein 121.12 64.45 155.13 1.28 0.28 4.60 4.32E-06 6.40E-03 disulfide bond formation protein B 93.44 39.16 126.00 1.69 0.37 4.52 6.26E-06 7.73E-03 MULTISPECIES: porin 383.67 545.81 286.39 -0.93 0.21 -4.46 8.06E-06 8.39E-03 Hsp70 family protein 23.04 42.13 11.59 -1.82 0.41 -4.44 9.06E-06 8.39E-03 type III glutamate--ammonia ligase 62.39 34.05 79.39 1.24 0.29 4.25 2.17E-05 1.78E-02 hydrogenase expression protein, partial 28.18 55.62 11.71 -2.32 0.57 -4.04 5.39E-05 3.63E-02 pilus assembly protein 22.96 7.02 32.53 2.23 0.55 4.05 5.11E-05 3.63E-02 elongation factor G, partial 234.31 114.09 306.44 1.42 0.36 4.00 6.32E-05 3.90E-02

Table 8: Functions subsystems using SEED that were significantly different in abundance between the pre and post gemmule sponge samples. Differential abundance was calculated using DESeq2 and the table includes mean (X), log2 fold change, indicating if the subsystems increased or decreased in abundance, the error associated with the fold change (lfcSE), as well as the statistic (stat), p-value, and adjusted p-value.

Function Subsystems baseMean PreMean PostMean log2Fold lfcSE stat pvalue padj Change Motility and Chemotaxis 9664.44 8456.79 10389.02 -0.30 0.13 -2.32 2.04E-02 5.72E-01 OTHER 33854692.19 30797651.1 35688916.85 -0.21 0.29 -0.74 4.59E-01 9.71E-01 Respiration 787010.00 804586.93 776463.84 0.051 0.10 0.54 5.90E-01 9.71E-01 Cofactors, Vitamins, Prosthetic Groups, Pigments 74716.78 76855.51 73433.54 0.07 0.06 1.08 2.82E-01 9.71E-01 Cell Wall and Capsule 20212.54 19408.16 20695.16 -0.09 0.10 -0.98 3.28E-01 9.71E-01 Virulence 10933.64 11213.52 10765.72 0.06 0.12 0.48 6.35E-01 9.71E-01 Photosynthesis 6732.09 7199.09 6451.89 0.16 0.30 0.53 5.99E-01 9.71E-01 Phages, Prophages, Transposable Elements, Plasmids 4550.34 4270.72 4718.11 -0.14 0.21 -0.69 4.90E-01 9.71E-01 Sulfur Metabolism 2268.71 2168.71 2328.71 -0.10 0.14 -0.73 4.65E-01 9.71E-01 Iron Acquisition and Metabolism 1107.96 1143.57 1086.59 0.07 0.12 0.62 5.38E-01 9.71E-01 Phages, Prophages, Transposable Elements 880.20 707.66 983.73 -0.48 0.72 -0.66 5.08E-01 9.71E-01

Dormancy and Sporulation 70.00 31.07 93.35 -1.59 0.93 -1.71 8.80E-02 9.71E-01

Table 9: Functions that were significantly different in abundance between the pre and post gemmule

sponge samples from two genera (Polynucleobacter and Flavobacterium) that were significantly different

and abundant in Table 5. Differential abundance was calculated using DESeq2 and the table includes

mean (X), log2 fold change, indicating if the functions increased or decreased in abundance, the error

associated with the fold change (lfcSE), as well as the statistic (stat), p-value, and adjusted p-value.

Organism: Polynucleobacter

Function BaseMean PreMean PostMean log2FoldChange lfcSE stat pvalue padj magnesium chelatase subunit H 29.45 10.74 40.67 1.88 0.34 5.55 2.87E-08 2.12E-05 light-harvesting protein 479.88 229.57 630.06 1.46 0.30 4.86 1.16E-06 4.27E-04 chlorophyllide a reductase iron protein subunit X 18.20 8.66 23.92 1.52 0.42 3.62 2.91E-04 3.58E-02 hydroxymethylglutaryl-CoA lyase 21.88 36.08 13.36 -1.42 0.38 -3.77 1.62E-04 3.58E-02 photosynthetic reaction center subunit L 18.63 8.03 25.00 1.58 0.43 3.64 2.72E-04 3.58E-02 photosynthetic reaction center subunit M 28.84 14.21 37.62 1.40 0.38 3.68 2.36E-04 3.58E-02 transaldolase 59.41 37.27 72.69 0.93 0.26 3.58 3.43E-04 3.62E-02 Organism: Flavobacterium

Function BaseMean PreMean PostMean log2FoldChange lfcSE stat pvalue padj fibronectin type III domain-containing protein, partial 66.18 132.94 26.13 -2.30 0.42 -5.51 3.65E-08 9.42E-06 acyl-CoA desaturase 113.24 166.69 81.17 -1.07 0.21 -5.04 4.62E-07 5.96E-05 L-threonine 3-dehydrogenase 182.48 266.18 132.27 -1.04 0.23 -4.55 5.24E-06 4.51E-04 hypothetical protein 7982.82 9801.24 6891.77 -0.51 0.12 -4.34 1.43E-05 9.22E-04 4-hydroxyphenylpyruvate dioxygenase 49.30 72.15 35.60 -1.04 0.25 -4.22 2.43E-05 1.25E-03 NADH-quinone oxidoreductase subunit N 1868.23 2685.51 1377.87 -0.97 0.23 -4.17 2.99E-05 1.29E-03 beta-glucosidase 44.23 66.20 31.05 -1.12 0.28 -4.08 4.53E-05 1.67E-03 alkaline phosphatase family protein 110.27 150.17 86.33 -0.82 0.21 -4.00 6.32E-05 2.04E-03 fumarylacetoacetase 121.12 172.37 90.37 -0.96 0.24 -3.95 7.84E-05 2.25E-03 DUF11 domain-containing protein 25.25 42.17 15.10 -1.49 0.38 -3.91 9.31E-05 2.40E-03 proton-translocating NADH-quinone oxidoreductase chain N 134.00 196.30 96.63 -1.06 0.28 -3.85 1.19E-04 2.79E-03 DUF3347 domain-containing protein 26.76 41.67 17.82 -1.23 0.33 -3.74 1.87E-04 4.03E-03 cell division protein 575.06 755.24 466.96 -0.70 0.19 -3.63 2.80E-04 5.15E-03 RagB/SusD family nutrient uptake outer membrane protein 256.25 340.05 205.98 -0.72 0.20 -3.65 2.61E-04 5.15E-03 aconitate hydratase 147.85 204.82 113.66 -0.88 0.24 -3.61 3.05E-04 5.25E-03 MULTISPECIES: hypothetical protein 294.77 390.34 237.42 -0.73 0.21 -3.43 6.06E-04 9.77E-03 MULTISPECIES: ornithine--oxo-acid transaminase 309.37 402.29 253.62 -0.68 0.20 -3.36 7.82E-04 1.19E-02 ATP:cob(I)alamin adenosyltransferase 54.99 78.03 41.16 -0.95 0.29 -3.25 1.15E-03 1.63E-02 alpha-glucosidase 31.97 47.01 22.94 -1.02 0.32 -3.24 1.20E-03 1.63E-02 beta-mannosidase 39.35 56.16 29.27 -1.00 0.31 -3.20 1.38E-03 1.78E-02 RNA-binding protein 225.23 292.08 185.13 -0.66 0.21 -3.09 2.02E-03 2.48E-02 aminotransferase 22.55 33.12 16.21 -1.08 0.35 -3.04 2.37E-03 2.78E-02 small GTP-binding protein, Ras family 22.60 33.53 16.05 -1.06 0.36 -2.98 2.89E-03 3.25E-02 ankyrin repeat domain-containing protein 45.18 61.73 35.25 -0.85 0.29 -2.96 3.10E-03 3.33E-02 T9SS C-terminal target domain-containing protein 1199.97 1401.43 1079.09 -0.37 0.13 -2.82 4.84E-03 5.00E-02

Appendix 1. Sponge Germination Protocol Overview: Gemmules collected from parental sponge in the field have the ability to be hatched in the laboratory. In order to do so they must be isolated and washed so they are in pristine condition upon storage in 4oC. Gemmules are typically stored for an extended period of time, prior to hatching. Gemmules take about 4 to 7 days to start hatching and attaching themselves to the substrate in the well plates. They typically continue to grow in size for 7-10 days before starting to desiccate. Collection: • Collect sponge tissue with visible gemmules with a sterile razor blade • Tissue is then stored in 50 mL falcon conical tube of filtered river water or cold Strekal’s media covered with foil • Place tube containing sample on ice until returned to campus • Store in fridge (4oC) still covered with foil to block out light for a year or more Strekal’s Medium: • An aqueous solution enabling gemmules attachment (Strekal and McDiffett, 1974). • 10X Strekal’s Medium contains 0.9mM of MgSO4-7H2O, 0.5mM CaCO3, 0.1 mM Na2SiO3-9H2O and 0.1 mM KCl • MgSO4-7H2O 246.47푔 1 푚표푙 0.9 푚푚표푙 o ∗ ∗ ∗ 1퐿 = 0.222푔 푚표푙 1000 푚푚표푙 퐿 • CaCO3 * 2H2O 100.086푔 1 푚표푙 0.5 푚푚표푙 o ∗ ∗ ∗ 1퐿 = 0.05푔 푚표푙 1000 푚푚표푙 퐿 • Na2SiO3-9H2O 284.20 푔 1 푚표푙 0.1 푚푚표푙 o ∗ ∗ ∗ 1퐿 = 0.028푔 푚표푙 1000 푚푚표푙 퐿 • KCl 74.5513 푔 1 푚표푙 0.1 푚푚표푙 o ∗ ∗ ∗ 1퐿 = 0.007푔 푚표푙 1000 푚푚표푙 퐿

Separating Gemmules From Maternal Tissue: • Most steps were done following the ‘Pick, Wash and Hatch Gemmules Handbook’ • Important to note all following steps should be done on ice and keeping the Strekal’s Medium cold during the procedure with minimum light exposure (I dimmed the room as the dissecting microscope produced plenty of light) • Take a small portion of the sponge tissue and place it into a small petri dish containing cold Strekal’s Medium • Using a dissecting microscope and teasing needles, carefully removal all maternal tissue from the gemmules. o Be careful not to pop the gemmules-they won’t be useful afterwards • A disposable pipette can be used to transfer clean gemmules into a new falcon conical tube of Strekal’s solution on ice to insure they don’t get too hot by the microscope light • Left over maternal tissue can either be disposed of or stored for later use

• Continue to repeat this process until all maternal tissue has been sorted through and all gemmules removed • (Gemmules may be stored at this stage if needed-be sure to label as unwashed if you do) Washing Gemmules: • Prepare a hydrogen peroxide solution in a 15mL Falcon conical tube by diluting 400 L of 30% hydrogen peroxide in 13.6 mL (Handbook says 500L in 13mL- but upon email the author says she uses a lower percentage to wash the gemmules) • Gather all the clean gemmules into about 1mL of Strekal’s solution (by dumping and pipetting out excess solution) and pipette the combination into the hydrogen peroxide solution tube o Adding the 1mL of Strekal will permit the hydrogen peroxide to reach a concentration of 1% (or less) • Invert the 15mL tube every minute for a duration of 3 minutes o This will permit removal of algae and to weaken gemmules membrane o Little bubbles should appear • Help the gemmules settle to the bottom of the tube by softly tapping the tube o Some of them will not settle because of the bubbles, don’t worry and continue • Remove the used solution by inverting the tube above a petri dish and rapidly pick up the few gemmules from the dish and put them back into the tube • Fill the tube will new cold Strekal’s medium and invert it immediately several times • Change the solution again using the same inverting method and replace with new medium • Incubate gemmules for 3 minutes in the new solution, inverting the tube every minute o All gemmules which are still floating can be removed at this step o As they are empty and contain no cells in them • Repeat the last step 6 to 8 more times o You may do more if you feel it’s needed- but do no less as you want to be sure all the peroxide is removed from the gemmule coat • After that the gemmules are washed and ready to use. They can also be stored at 4oC at this time for later use o Please keep in mind 4oC is the fridge, not the freezer Hatch Gemmules: • Remove gemmules and Strekal’s from fridge and reach room temperature o May proceed to next steps as this takes awhile • Open an 8-well plate and put 1 large or 2 small cover glasses in each well • Add 1-2 mL of Strekal’s Medium in each well (or until the cover glass is covered) • Remove any air between the well and cover-glass with a pipette tip • Once gemmules and Strekal’s reach room temperature, place 3-4 gemmules in each well

o Try to make sure the gemmules don’t group together so they don’t grow into one another • Store gemmules in a dark, room temperature place for 2-3 days without disrupting them o I used a labeled cabinet so there was no risk of them being exposed to light in theses beginning stages • Handbook says to change the solution every 24 hours o I never changed the solution (afraid to suck up the gemmules or disrupt their attachment) and they still continued to grow

Vita

Taylor Ann Strope was born in Knob Noster, Missouri to Anthony and Tracy Strope.

She was raised in Hamilton, Illinois by Anthony Strope. She graduated from Nevada High

School in Missouri in 2014. The following fall semester, she entered at Texas A&M

University of Galveston to study Marine Biology and Marine Fisheries, and in May 2018 she was awarded the Bachelor of Science degree in dual majors. In the fall of 2018, she continued progress toward a Master of Science degree at the Appalachian State University in

Biology with a concentration in Cell and Molecular. The M.S. was awarded in August 2020.

Following the M.S. degree Taylor Strope will be continuing on to receive her Ph.D. in the

Interdisciplinary Graduate Program in Biomedical Sciences at University of Kansas Medical

Center in Kansas City, Kansas.